{"id":"ffd0b6d3-17e8-4d83-b7a7-896c23c347c6","arxiv_id":"2608.09896","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"The first complete census of dust mass, disk size, and CO-based gas mass for Taurus Class 0/I protostellar disks finds median dust masses of 5.5 Earth masses at 345 GHz and 39 Earth masses at 33 GHz, and a broad gas-to-dust ratio distribution.","lead":"Astronomers measured dust and gas in 26 of the youngest planet-forming disks in the nearby Taurus star-forming region, using ALMA and VLA radio telescopes. They found these embedded disks are generally small and low-mass, with wide variation in their gas-to-dust ratios.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The two continuum mass scales disagree by a factor ~5.7; until that offset is resolved, the headline 5.5 Mearth and 39 Mearth medians and the gas-to-dust ratios built on them are not robust.","rationale":"The reader's weakest assumption was the optically thin isothermal single-opacity conversion (Eq. 1). I agree that this is the foundational premise, but I want to sharpen it: the paper's own data contain a factor-5.7 discrepancy between the 345 GHz and 33 GHz mass scales, so the failure of the assumption is not merely an external uncertainty - it is an internal contradiction in the reported medians. This makes the central quantitative claims conditional in a concrete, testable way. I do not think this changes the verdict: the reader already recommended CONDITIONAL, and my concern reinforces that conclusion. I set agreement_with_reader to 'partial' because my focus is on the internal 5.7 offset rather than on the general opacity/temperature systematics the reader emphasized. A dedicated radiative-transfer reconciliation of the two bands is the cleanest single check, since it would directly tell us whether the true mass scale is near 5.5 Mearth, near 39 Mearth, or somewhere in between.","tokens_in":40887,"tokens_out":8989,"duration_ms":94306,"concrete_test":"Run source-by-source radiative transfer fits (pdspy/RADMC-3D, as in Sheehan et al. 2022) to the combined ALMA 345 GHz and VLA 33 GHz visibility data, allowing optical depth, Tdust(r), and a free-free component at 33 GHz to vary self-consistently, and compare the resulting 345 GHz and 33 GHz dust masses. If the best-fit masses converge to within ~30% and the converged mass scale differs by more than a factor of 2 from either reported median, the corresponding headline median and all gas-to-dust ratios built on it are biased by that factor. Repeating the same fits for the Taurus Class II comparison sample would also test whether the 'comparable masses at 345 GHz' conclusion survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing concern is internal: the paper applies the same optically thin, single-opacity mass formula (Eq. 1, Section 3.1) at 345 GHz and 33 GHz and obtains median dust masses of 5.5 Mearth and 39 Mearth for the same sources, then states that the ALMA-based masses would need to be multiplied by ~5.7 to match the VLA distribution. Under the adopted opacity prescription, the two bands should give the same physical mass to within the assumed κν scaling, so a factor-5.7 offset means at least one of the two reported median dust masses is not a true dust mass. The possible resolutions are not benign: optically thick 345 GHz emission (the paper's own estimate is that τ~1-2 biases masses low by factors 1.6-2.3, and more for higher τ) and/or free-free contamination at 33 GHz (median ~10-15%, with source-level contributions up to ~100%) would move the true median substantially. The headline 5.5 Mearth value, the 39 Mearth value, and every gas-to-dust ratio computed from the 345 GHz masses inherit this unresolved factor. The paper acknowledges the systematics, but it presents both frequency-specific medians as measured quantities, leaving the central numerical claims model-dependent on an internal inconsistency that could shift the fiducial mass scale by up to ~5.7.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents ALMA Band 7 (345 GHz) and VLA Ka-band (33 GHz) continuum observations, together with 13CO and C18O J=3–2 line data, for 26 Taurus Class 0/I protostellar systems (35 individual disks). The authors measure dust masses and radii from the continuum using the standard optically thin formula (Eq. 1), report median values of 5.5 M⊕ and 28 AU at 345 GHz and 39 M⊕ and 32 AU at 33 GHz, and infer gas masses by matching Keplerian-masked CO isotopologue luminosities to a radiative-transfer model grid, yielding a median gas mass of 6.7e-4 M⊙ and a median gas-to-dust ratio of 26. They compare the Taurus Class I disks with samples in Orion, Perseus, and Ophiuchus, and with Taurus Class II disks, concluding that Taurus occupies an intermediate position in flux and mass and that Class 0/I disks are brighter but not more massive than Class II at 345 GHz.","tokens_in":41334,"tokens_out":10089,"duration_ms":87044,"significance":"If the quantitative results hold, this would be the first region-complete census of dust masses, sizes, and gas masses for embedded disks in Taurus, providing important constraints on early disk evolution and on the gas-to-dust ratio at the Class 0/I stage. The paper's strengths include a uniformly selected sample, joint ALMA/VLA analysis, careful Kaplan–Meier treatment of upper limits, and explicit sensitivity tests for CO abundance and freeze-out temperature. However, the central numerical claims are weakened by an unresolved factor ~5.7 discrepancy between the two continuum-derived mass scales and by gas-mass estimates that inherit the adopted Keplerian-mask and stellar-mass assumptions. These issues must be addressed before the reported medians can be accepted as robust census values.","major_comments":[{"comment":"The paper reports median dust masses of 5.5 M⊕ at 345 GHz and 39 M⊕ at 33 GHz for the same sources, using the same opacity law and the same optically thin formula (Eq. 1), and notes that the ALMA-based masses would need to be multiplied by about 5.7 to match the VLA-based distribution. Under the adopted κν ∝ ν scaling, the two bands should yield the same physical mass, so this offset implies that at least one of the two mass scales is biased. The paper's own estimates of the optical-depth bias (factors 1.6–2.3 for τ ~ 1–2) and free-free contamination at 33 GHz (median 10–15%, up to ~100%) are of comparable magnitude. Because both medians are quoted in the Abstract and used in the cross-region and Class II comparisons, the central quantitative claims are not robust until this discrepancy is resolved, either by applying corrections or by propagating the systematics into the reported medians and confidence intervals.","section":"§3.1, Eq. (1), and the discussion following Table 4"},{"comment":"The stellar masses adopted for the Keplerian masks are tuned by eye to match the observed 13CO line extents (Section 3.2), and the model grid used to derive gas masses sets its stellar-mass range from these same Keplerian-mask estimates (Section 3.3). The measured line fluxes, and hence the inferred gas masses, are therefore not statistically independent of the assumed stellar masses and mask geometry. The reported gas-mass uncertainties in Table 5 are 16th–84th percentiles of the match-weighted model distribution and do not include uncertainties in the adopted M*, inclination, position angle, z/r, or rout values. The consistency check with radiative-transfer stellar masses for a subset is encouraging, but a quantitative sensitivity analysis of the inferred gas masses to plausible variations in the mask parameters is needed to support the median gas mass of 6.7e-4 M⊙ and the resulting gas-to-dust ratios.","section":"§3.2 and §3.3"},{"comment":"The gas-to-dust ratios are computed using the 345 GHz dust masses from Eq. (1), which the paper itself notes are biased low by optically thick emission (Section 3.1, factors 1.6–2.3 for τ~1–2 and more for higher τ). Because the CO-inferred gas masses are not corrected for this bias, the reported median gas-to-dust ratio of 26 and the 16th–84th percentile range of 8–147 are systematically too high. This propagation should be either quantified and corrected, or the gas-to-dust ratio distribution should be presented with this caveat as the dominant uncertainty rather than as a measured range.","section":"§3.3 and Figure 9"},{"comment":"The statement that Taurus Class 0/I disks are comparable in 345 GHz dust mass to Taurus Class II disks (medians 5.5 vs 5.6 M⊕) is used to argue that embedded disks are not more massive than Class II disks. If the Class 0/I disks are partially optically thick while the Class II sample is not, this equality could be an artifact of the bias discussed in Section 3.1. The paper discusses this possibility in Section 4.2, but the Abstract and Section 5 present the conclusion without that caveat. A quantitative test (e.g., comparing spectral indices or applying an optical-depth correction) is needed before this conclusion can be stated as a primary result.","section":"§4.2 and Abstract"}],"minor_comments":[{"comment":"The sentence 'our ALMA datasets includ both 12 m and ACA 7 m coverage' contains a typo; 'includ' should be 'include'.","section":"§2.2"},{"comment":"Please verify the opacity power-law exponent in the formula κν = 10 (ν/1000 GHz) cm2 g−1; the stated values κ345 = 3.45 and κ33 = 0.33 correspond to κν = 10(ν/1000 GHz) cm2/g with exponent unity, so the printed formula should be checked for a missing or misplaced exponent.","section":"§3.1"},{"comment":"The phrase 'outliers fall below 10^3 in luminosity' is unclear; if the intended threshold is 10^-3, as the surrounding text suggests, the exponent should be corrected.","section":"Figure 4 caption"},{"comment":"The entry for 04264+2433AB lists two systemic velocities and two stellar masses; please add an explanatory note clarifying which value applies to which component or stating that the mask was constructed for the combined system.","section":"Table 5"},{"comment":"The description of the VANDAM–Perseus 9 mm flux as 'averaging the two bands' could introduce a bias if the spectral index varies across the sample; a short justification or a consistency check using the measured spectral indices would improve clarity.","section":"§4.1"}],"recommendation":"major_revision","confidential_remarks":"The paper presents a valuable observational dataset and is transparent about many systematics. My main concern is that the factor ~5.7 offset between the two continuum-derived mass scales is left unresolved while both scales are used in headline results; this is a load-bearing internal inconsistency that cannot be fixed by caveats alone. The gas-mass method also needs to demonstrate robustness to the mask and stellar-mass choices. I believe a major revision addressing these points is appropriate; the data and comparisons will be useful to the community."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on the NEST II paper. What you should know first: it delivers the first region-complete Taurus Class 0/I disk census with both dust continuum and CO-based gas masses, and it is careful and transparent about its systematic uncertainties. But the same optically thin, single-opacity formula applied to the ALMA 345 GHz and VLA 33 GHz data gives median dust masses of 5.5 Mearth and 39 Mearth for the same sources, a factor ~5.7 offset that the paper explicitly acknowledges but does not resolve. Both medians go into the abstract as headline values, and the gas-to-dust ratios use the 345 GHz scale. So the central numbers are model-dependent in a way that could shift them by a factor of several.\n\nWhat's genuinely new: the region-complete sample, the frequency-matched cross-region comparisons (Taurus vs Orion/Ophiuchus/Perseus at 345 and 33 GHz separately), and the first gas-to-dust distribution for Taurus Class 0/I from CO radiative-transfer modeling. The paper ships the source tables and is unusually candid about what is not known: optical depth biases (factors 1.6-2.3 at tau 1-2), opacity prescription scatter (order of magnitude), free-free at 33 GHz (median ~10-15%, but up to 100% in some sources), and CO abundance sensitivity (factors ~1.4-10 from x_CO 1e-5 to 1e-3). The Kaplan-Meier statistics with upper limits are appropriate.\n\nThe soft spot is proportionally the same one the stress test flags: the 5.7 factor between the two mass scales is load-bearing. The paper's own caveat that relative dust-mass trends should be read as \"suggestive rather than exact\" appears in Section 3.1, after the abstract has already presented 5.5 and 39 Mearth as measured medians. A serious referee should push for either a combined mass scale, a clear fiducial choice with justification, or a restructured presentation that demotes the absolute masses to model-dependent values. The gas masses are also grid fits with fixed ISM abundances, not independent measurements; they are best treated as a first pass.\n\nWho benefits: observers studying embedded disks and planet formation initial conditions, and anyone comparing disk properties across star-forming regions. It deserves a serious referee and likely publication after revision. I would cite it for the census and the cross-region CDFs, and I would bring it to a reading group to discuss the mass-scale systematics.","headline":"Useful, honest census of Taurus Class 0/I disks, but the two continuum-derived dust mass scales disagree by ~5.7 and the paper leaves the central medians model-dependent.","tokens_in":41809,"tokens_out":3320,"would_cite":true,"duration_ms":32518,"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":"The first region-complete census of Taurus's 26 embedded disks reports median dust masses of 5.5 Earth masses at 345 GHz and 39 Earth masses at 33 GHz, with CO-based gas-to-dust ratios centered near 26.","keywords":["protostars","protoplanetary disks","Class 0/I disks","dust mass","gas-to-dust ratio","Taurus molecular cloud","millimeter interferometry","disk radii"],"falsifier":"Fit each target's 345 GHz and 33 GHz continuum together with a longer-wavelength 1-3 cm measurement in a source-by-source radiative-transfer model that includes finite optical depth and free-free emission; if the corrected median dust masses fall outside the quoted 68% ranges (3.9-9.4 Earth masses at 345 GHz, 34-52 Earth masses at 33 GHz), or a C$^{17}$O-based gas census shifts the median gas-to-dust ratio far from 26, the headline numbers of this paper would need revision.","tokens_in":40668,"feed_emoji":"🌌","tokens_out":13467,"duration_ms":107987,"temperature":0.7,"pith_summary":"The paper sets out the first region-complete census of the earliest, envelope-embedded disks in Taurus: 26 Class 0/I protostellar systems observed with high-resolution 345 GHz and 33 GHz continuum imaging plus carbon monoxide isotopologue lines. Its central claim is that these nascent disks have median 345 GHz flux 71 mJy, dust mass 5.5 Earth masses, and radius 28 AU, and median 33 GHz flux 0.43 mJy, dust mass 39 Earth masses, and radius 32 AU, with a CO-inferred median gas mass of $6.7\\times10^{-4}$ solar masses. That gas mass translates into a median gas-to-dust ratio of 26 with a 16th-84th percentile range of 8-147, so many embedded disks hold far less gas per unit dust than the canonical interstellar ratio of 100. If correct, this census would mean that Taurus starts its disk lifetimes with dust reservoirs comparable to its older Class II disks at 0.87 mm, and it would place the region squarely between the larger, more massive disks of Orion and the smaller, less massive disks of Ophiuchus. The stakes are concrete: these are the initial solid and gas budgets from which planets must assemble.","feed_headline":"Youngest Taurus disks hold 5.5 Earth masses of dust","feed_subtitle":"First region-complete census of 26 embedded disks finds gas-to-dust ratios as low as 8, well below the interstellar 100.","key_machinery":"The argument runs on three linked machines. The first is a pair of interferometric continuum images, ALMA at 345 GHz and the VLA at 33 GHz, whose sub-arcsecond beams isolate a compact disk component from the surrounding envelope; every flux and size in the paper comes from two-dimensional Gaussian fits to those images. The second is the standard optically thin, isothermal dust-mass conversion, $M_{\\rm dust}=F_\\nu d^2/[\\kappa_\\nu B_\\nu(T_{\\rm dust})]$, with a luminosity-scaled dust temperature and a single power-law opacity at both frequencies; this is what turns raw flux into the headline Earth masses. The third is a Keplerian-masking plus radiative-transfer pipeline: $^{13}$CO and C$^{18}$O line cubes are masked to the velocity pattern expected from a rotating disk, and the masked line luminosities are matched against a disk-plus-envelope model grid to assign a gas mass. The radii come from deconvolved Gaussian major axes, and the censored medians are computed with a Kaplan-Meier estimator that carries upper limits.","core_discovery":"On the paper's own terms, the discovery is a set of population medians for the youngest disks in a single cloud, measured the same way for every source. At 345 GHz the sample gives a median flux density of 71 mJy (54-107 mJy at 68%), a median dust mass of 5.5 Earth masses (3.9-9.4), and a median radius of 28 AU (25-39); at 33 GHz the corresponding medians are 0.43 mJy (0.41-0.80), 39 Earth masses (34-52), and 32 AU (29-33). The CO-based gas masses have a median of $6.7\\times10^{-4}$ solar masses, giving gas-to-dust ratios with a mean of 147, a median of 26, and a 16th-84th percentile range of 8-147. The paper also claims that Taurus Class I disks are fainter and less massive than Orion Class 0/I disks, broadly comparable to Perseus Class I disks, and brighter and more massive than Ophiuchus Class I disks, while Taurus Class 0/I disks are brighter than Taurus Class II disks at both frequencies but not more massive at 345 GHz.","pith_inferences":["A natural extension the authors do not push: the factor-of-about-5.7 offset between the ALMA and VLA dust-mass scales is most simply explained by 345 GHz emission that is not perfectly optically thin, in which case the true dust masses lie between the two medians and the gas-to-dust ratios computed against 345 GHz masses are upper limits on the true ratios.","The same Keplerian-masking plus radiative-transfer pipeline could be run on the existing 345 GHz and 33 GHz data for Orion and Ophiuchus, giving a three-region census analyzed with identical assumptions; that would test whether the reported inter-region ordering survives a fully matched analysis.","A testable prediction follows from the low end of the gas-to-dust distribution: if ratios near 8 are real, those disks have already lost or sequestered most of their gas or grown their solids enough to lower the apparent dust opacity, and they should show accompanying spectral-index or CO-depletion signatures.","Because the Class II comparison adopts a cooler fixed dust temperature (25 K versus 43 K for Class 0/I), part of the 'comparable dust masses' result is sensitive to that temperature choice; recomputing both populations with a common luminosity-scaled temperature law is a cheap check."],"forward_implications":["If the 345 GHz median dust mass of 5.5 Earth masses is right, Taurus embedded disks start with about the same solid reservoir as the region's older Class II disks, so the steep Class 0/I-to-Class II dust-mass decline reported in cross-region studies is not universal.","If the CO-inferred gas masses are right, applying the canonical interstellar gas-to-dust ratio of 100 to Taurus embedded disks would overestimate their total gas by roughly a factor of four at the median.","If the 33 GHz dust masses are the more reliable tracer, the census implies roughly seven times more dust than the 345 GHz estimate, which changes the total-mass budget available for planet formation and pushes the inferred gas-to-dust ratios downward.","If the region ordering is real, disk assembly models must produce larger, more massive disks in Orion, intermediate ones in Taurus, and compact, low-mass disks in Ophiuchus from environmental differences alone, without invoking different stellar initial mass functions."],"supporting_citations":[{"why":"Supplies the optically thin flux-to-mass relation $M_{\\rm dust}=F_\\nu d^2/[\\kappa_\\nu B_\\nu]$ used to convert every continuum measurement into a dust mass.","marker":"Hildebrand 1983"},{"why":"Supplies the adopted opacity law $\\kappa_\\nu=10(\\nu/1000\\,{\\rm GHz})^2\\,{\\rm cm}^2\\,{\\rm g}^{-1}$, which sets the absolute dust-mass scale at both frequencies.","marker":"Beckwith & Sargent 1991"},{"why":"Provides the luminosity-scaled dust temperature prescription and the Orion Class 0/I comparison sample used for the cross-region censored distribution tests.","marker":"Tobin et al. 2020"},{"why":"Radiative-transfer models showing that compact continuum flux tracks disk emission, justifying the disk-envelope separation and informing the temperature treatment.","marker":"Sheehan et al. 2022"},{"why":"Outlines the CO line radiative-transfer approach for estimating gas masses and supplies the Taurus Class II gas-to-dust ratios used as the low-end comparison.","marker":"Williams & Best 2014"},{"why":"Defines the 26-system Taurus sample, the Class 0/I classification, and the adopted bolometric luminosities and temperatures.","marker":"Plante et al. 2026"},{"why":"Supplies the per-subregion Gaia distances used to scale fluxes to dust masses and radii and to normalize the comparison samples to a common distance.","marker":"Zucker et al. 2020"},{"why":"Provides the empirical centimeter-luminosity relations used to estimate the possible free-free contribution to the 33 GHz fluxes.","marker":"Tychoniec et al. 2018"}],"fun_headline_variants":["Taurus census: youngest disks hold 5.5 Earth masses of dust","Taurus disk gas-to-dust ratios run as low as 8","First complete census: Taurus gas-to-dust as low as 8","Youngest Taurus disks: median dust mass 5.5 Earths","Taurus disks: 5.5 Earth masses dust, gas-to-dust down to 8"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole mass census rests on assuming that a single opacity value, a luminosity-only temperature, and optically thin emission can convert each measured continuum flux into a dust mass, even though the paper itself notes that optical depths near 1-2 would raise the true masses by factors of 1.6-2.3, opacity choices vary by an order of magnitude, and the two frequencies disagree by a factor of about 5.7.","fun_headline_variants_meta":{"raw":{"variants":["Taurus census: youngest disks hold 5.5 Earth masses of dust","Taurus disk gas-to-dust ratios run as low as 8","First complete census: Taurus gas-to-dust as low as 8","Youngest Taurus disks: median dust mass 5.5 Earths","Taurus disks: 5.5 Earth masses dust, gas-to-dust down to 8"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00207,"raw_usage":{"total_tokens":8188,"prompt_tokens":1214,"completion_tokens":6974,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":830,"completion_tokens_details":{"reasoning_tokens":6871}},"tokens_in":830,"tokens_out":6974,"duration_ms":43468,"temperature":1.0,"reasoning_tokens":6871,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:42:10.908969+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fit each target's 345 GHz and 33 GHz continuum together with a longer-wavelength 1-3 cm measurement in a source-by-source radiative-transfer model that includes finite optical depth and free-free emission; if the corrected median dust masses fall outside the quoted 68% ranges (3.9-9.4 Earth masses at 345 GHz, 34-52 Earth masses at 33 GHz), or a C$^{17}$O-based gas census shifts the median gas-to-dust ratio far from 26, the headline numbers of this paper would need revision.","supporting_citations":[],"review_version":1}