{"id":"931aeec3-c5a3-4b5c-9b1e-8ef4aded9d0d","arxiv_id":"2507.03370","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A simulated Jupiter-mass planet's accretion heating creates a detectable CO bubble in the freeze-out zone of a protoplanetary disk, offering a new ALMA-based protoplanet detection channel.","lead":"The paper uses 3D radiation-hydrodynamic simulations to show that a luminous young planet can keep carbon monoxide from freezing out around it, creating a warm CO bubble that telescopes like ALMA could spot. If real, this gives astronomers a new, automated way to find forming planets in the cold outer parts of disks, where other detection methods often fail.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 9σ detection claim rests on an unpublished, unbenchmarked axial-flow subtraction pipeline (disckin) applied to idealized thermal-noise-only ALMA cubes; until it is released and validated on synthetic data, the central observability claim cannot be independently audited.","rationale":"I read the paper as a proof-of-concept: nominal 3D RHD simulation, radiative transfer, synthetic ALMA analysis. The physics chain—accretion heating creates a warm bubble, freeze-out contrast creates an emission spot—is coherent and supported by Appendices A and B. The reader's luminosity caveat is real and correctly weighted: S1 shrinks the bubble to 45% and S4 is undetectable. But in my judgment the single most proximate load-bearing link is the observability analysis itself. The 9-sigma number is not an observable; it is the output of an unreleased axisymmetric-fitting code whose benchmark is explicitly deferred. Because the bubble can be partially absorbed into the axisymmetric model (the paper says so), the detection significance is precisely the quantity most sensitive to the algorithm's details, median-filter parameters, and noise model. A conditional verdict is appropriate; the condition should be public validation of disckin and a recovery test on the synthetic cube, plus ideally the same test on S1. I therefore leave the reader's verdict unchanged.","tokens_in":23336,"tokens_out":7596,"duration_ms":101467,"concrete_test":"Release disckin (or reconstruct an axisymmetric background with an independent kinematic tool) and rerun the full Sect. 4.2 pipeline on the provided nominal synthetic C18O data cube, including realistic ALMA uv-plane sampling and calibration errors; require the candidate at the known planet position to remain at or above 5-sigma in at least one 0.2 km/s channel without retuning the median-filter kernel or the 3-sigma selection. If the peak drops below ~5-sigma or appears at the wrong position, the 9-sigma claim is not robust enough to support automatic detections.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's load-bearing quantitative claim is the 'peak PS residual ... at 9 sigma at +1.6 km/s' after axial-flow subtraction (Sect. 4.2). This number is produced by disckin, which the paper identifies as 'Casassus et al. in prep' and whose 'structure and physical conditions ... will be benchmarked against the input hydrodynamics in an upcoming article.' That is a self-declared missing validation. The subtraction also involves a median filter with a 5-beam kernel and a 3-sigma pixel selection whose robustness is not demonstrated, and the synthetic observations deliberately 'neglected systematic errors due to sparse uv-plane sampling' and assume perfect calibration, so the residuals are thermal-noise-limited only. The paper itself notes that at disk velocities the disckin model fits part of the bubble into the axisymmetric background, so the surviving 9-sigma signature is exactly the part of the signal least protected against algorithmic over-subtraction. If an independent implementation or an actual release of disckin removes the bubble into the background at all channels, or if realistic uv/calibration noise reduces the peak below point-source detection thresholds, the 'automatic detection' headline fails even if the luminosity caveat is set aside.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents 3D two-fluid radiation-hydrodynamic simulations (FARGO3D with a two-population dust model and frequency-averaged radiation transport) of a protoplanetary disk hosting a Jupiter-mass planet at 120 au, embedded in the CO freeze-out region beyond the snowline. The planet's accretion luminosity, parameterized by a mass-doubling time t_acc = 0.1 Myr (L_p = 2.8e-3 L_sun, T_p,e ~ 4000 K), maintains a warm bubble with T > 21 K extending slightly beyond the Hill sphere, so gas-phase CO survives locally despite the cold background. Post-processing with RADMC-3D and an instantaneous freeze-out prescription (ffreeze = 1e-5) yields synthetic C18O(2-1) channel maps in which the bubble appears as a low-intensity spot between the dragonfly wings of the line-forming surfaces. After adding thermal noise and subtracting an axially symmetric kinematic disk model computed with the unpublished disckin package, the point-source residual peaks at 9 sigma at +1.6 km/s. A companion result is that large dust grains are evacuated from the circumplanetary region by the horseshoe flow, making the planet nearly invisible in 1 mm continuum, consistent with the scarcity of ALMA circumplanetary detections. Appendix B reports sensitivity runs: a fivefold lower luminosity (S1) shrinks the bubble to 45% of the nominal diameter, an altered opacity law (S2) partly compensates, a planet at 85 au (S3) gives a similar bubble, and a 20 M_Earth planet (S4) produces a bubble that is not detectable.","tokens_in":23658,"tokens_out":17283,"duration_ms":205218,"significance":"Taken at face value, the paper offers a new, falsifiable observable channel for embedded accreting protoplanets outside the CO snowline: a thermally revived CO bubble that is visible even when the circumplanetary continuum is not. The mechanism is an emergent property of the forward RHD model rather than a fit to any target observation, so the circularity concern does not land; the audibility concern does, because the quantitative 9 sigma claim rests on a subtraction pipeline whose validation is deferred and on thermal-noise-only synthetic data. The paper deserves credit for its transparency: the parameter set is fully tabulated, the simulation codes are public, the noise calibration and median-filter steps are described in enough detail to reproduce, and the Appendix B sensitivity suite brackets the detectability in luminosity and planet mass. The continuum-depletion result (horseshoe-flow removal of large grains) is a clean mechanistic prediction that directly addresses a known observational tension.","major_comments":[{"comment":"The abstract's quantitative headline (\"the peak PS residual is at 9 sigma at +1.6 km/s\") is produced by the disckin package, cited as \"Casassus et al. in prep\", whose underlying model \"will be benchmarked against the input hydrodynamics in an upcoming article.\" The manuscript itself reports that at disk velocities the disckin model fits part of the bubble into the axisymmetric background, so the surviving residual is exactly the portion that the model cannot absorb. Since the subtraction depends on tunable choices (the free parameter Rfreeze, the 5-beam median kernel, and the 3 sigma pixel selection), the reported significance cannot be independently audited until either a full algorithmic description of disckin is supplied or a benchmark is shown, for example a null test on a planet-free control simulation or a comparison of the disckin model surfaces with the actual hydrodynamics. I request that this validation be added to the paper, or that the detection claim in the abstract be demoted to the raw channel-map identification of Fig. 5a, which does not depend on disckin.","section":"§4.2, disckin subtraction and the 9 sigma claim"},{"comment":"The authors state in §5 that \"the accretion luminosity, which is of critical importance for the formation of the bubble, is parametric and was not obtained in a self-consistent manner.\" The sensitivity runs quantify the stakes: with L_p reduced by a factor of about 5 (S1), the bubble shrinks to 45% of its nominal diameter, and a 20 M_Earth planet with the same t_acc (S4) gives a bubble of 30% nominal size that \"cannot be localized as a significant point source residual\" even in a ~7 d integration. Because the true accretion luminosities of embedded planets at ~100 au span a wide and poorly constrained range, the observability claim holds only for a subset of that range. The abstract's unconditional phrasing (\"enabling new automatic detections of forming protoplanets\") should therefore be qualified by the luminosity and planet-mass conditions established in Appendix B.","section":"§5 and Appendix B, parametric luminosity"},{"comment":"The 9 sigma significance refers to a peak pixel in residuals that are thermal-noise-limited by construction: the text states that \"systematic errors due to sparse uv-plane sampling\" are neglected and perfect calibration is assumed. Real ALMA data contain correlated nonthermal residuals, and the 5-beam median filter correlates the noise in the residual maps, so a peak-over-rms ratio is not a matched-filter detection significance. I ask the authors to report a PSF-weighted or matched-filter S/N for the bubble and to state the number of independent beams searched, so that a trials-corrected significance can be evaluated; this can be done with the existing residual cubes and does not require new hydrodynamics.","section":"§4.2, synthetic observation setup and detection statistics"},{"comment":"The nondetection around a 20 M_Earth planet rests on a main-stage run of only 25 orbits (Table B.1), i.e., about 20 kyr at 85 au, during which large dust is artificially removed from the planet's vicinity to keep the dust-to-gas ratio below 90%. It is not demonstrated that the bubble structure has converged on this timescale or that the dust-removal procedure does not alter the local thermal balance. A longer integration or an explicit convergence check for the S4 bubble would make the conclusion that subthermal-mass planets are undetectable more robust.","section":"§B.3, simulation S4"}],"minor_comments":[{"comment":"The freeze-out treatment (T <= 21 K, ffreeze = 1e-5) and the neglect of photodissociation and microturbulence are acknowledged, but a sentence assessing their effect on the bubble contrast specifically (as distinct from the dragonfly wings) would help; the claim that the bubble is optically thick suggests the intensity is set by temperature rather than by ffreeze, and this robustness could be stated explicitly.","section":"§2.4"},{"comment":"The phrase \"very rich observable chemistry\" goes beyond what is modeled: the simulations cover only CO and its isotopologs with an identical LTE excitation treatment. Suggest rephrasing, for example, \"the bubble is visible in multiple CO isotopologs.\"","section":"§Abstract"},{"comment":"The caption's \"Symmetric velocity channels\" should be defined; presumably these are channels symmetric about the systemic velocity (v and -v), which is not stated.","section":"§4.2 and Fig. 6"},{"comment":"The description of the noise calibration (rms \"amplified by a factor close to sqrt(N), where N is the number of pixels in the solid angle covered by the Gaussian beam\") is hard to follow; please state directly what noise level applies to the beam-smoothed cubes, and the assumed integration time and array configuration.","section":"§4.2"},{"comment":"The convergence information is given only for Sigma_g (change of 5% per 100 orbits); a corresponding check on the size of the T = 21 K bubble (for example, its azimuthally averaged extent during the last few hundred orbits) would strengthen the nominal case.","section":"§3.1 and Fig. 4"},{"comment":"The minimum dust density of 10^{-180} g cm^{-3} mentioned in the discussion of the high-altitude damping zones is a striking numerical value; please state the actual floor value imposed in the code and confirm that the midplane dust evolution is unaffected by it.","section":"§2.3.3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is transparent about both main contingencies (parametric luminosity and the unvalidated disckin pipeline), and the Appendix B sensitivity suite is a real strength. The editor may wish to consider whether the abstract's '9 sigma' claim should be published before the promised disckin benchmark appears; as written, that number cannot be independently checked, even by a specialist, without access to the pipeline. Since the disckin developer is a coauthor and the validation is described as coming in an upcoming article, I treat this as a fixable reporting gap rather than a fundamental flaw. The paper fits A&A's scope as a model-prediction paper. One further scope comment: the abstract's claims of 'very rich observable chemistry' and 'new automatic detections' are broader than the evidence currently supports, and the revision should align the abstract with the paper's own caveats."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Chrenko et al. run a careful 3D radiation-hydrodynamic simulation of a Jupiter-mass planet at 120 au, outside the CO snowline, and show that accretion heating sustains a warm gas-phase CO bubble in the frozen-out void. The bubble shows up in synthetic C18O channel maps, and they make a plausible case that this could be an automated detection channel for embedded protoplanets that leave no continuum and no velocity kinks. The physics is sound within the stated assumptions: two-stage relaxations, 800 orbits, a resolved Hill sphere, two-population dust opacities, and honest sensitivity runs in Appendix B. The dust-depletion mechanism via the horseshoe flow, explaining why CPDs are rarely seen in ALMA continuum, is a valuable and falsifiable addition.\n\nThe soft spots are real, though. The 9-sigma residual after axial-flow subtraction depends on disckin, which is unpublished and unbenchmarked, and the synthetic observations deliberately neglect uv-sampling and calibration errors. Until disckin is released or validated on synthetic data, the central detection claim can't be independently audited. The paper itself concedes that at disk velocities the disckin model fits part of the bubble into the axisymmetric background, so the surviving 9-sigma signal is exactly the part most vulnerable to over-subtraction. And the entire bubble hinges on the parametric accretion luminosity Lp = 2.8e-3 Lsun; cutting it by a factor of 5 shrinks the bubble to 45% extent, and a 20-Earth-mass planet produces a bubble that is not detectable. So the 'automatic detections' headline is conditional on a favorable combination of luminosity, opacity, inclination, and isotopolog.\n\nNone of this kills the paper. It is a proof of concept, and it is written like one. The authors are explicit about the parametric luminosity and the lack of chemistry. The synthetic observation procedure is described in enough detail to be reproduced once disckin is available. I'd send it to a serious referee; the referee should push for a disckin release or a benchmark against the input hydrodynamics, and should ask for a pilot search in existing ALMA archives as a first test.\n\nWho is this for? Planet formation theorists and ALMA observers working on protoplanet detection. It's worth discussing in a reading group even if the central claim is not yet fully auditable.","headline":"A solid simulation-based proof of concept for a new protoplanet detection channel, with the headline 9-sigma claim resting on an unpublished subtraction pipeline and a parametric luminosity.","tokens_in":24138,"tokens_out":3051,"would_cite":true,"duration_ms":32037,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A luminous accreting protoplanet outside the CO snowline keeps a warm bubble of gas-phase CO around itself, and the bubble shows up as a 9-sigma residual in synthetic ALMA C18O channel maps.","keywords":["protoplanetary disks","CO freeze-out","accretion luminosity","protoplanet detection","circumplanetary disk","planet-disk interactions","radiative transfer","hydrodynamics"],"falsifier":"Look for the bubble in a real disk with an independently confirmed accreting giant planet near 100 au: a C18O(2-1) observation at about 70 mas resolution, 0.2 km/s channels, and a deep integration should show a point-source residual above roughly 6$\\sigma$ at the planet's position after subtraction of the axisymmetric disk flow. A clean non-detection that cannot be blamed on viewing geometry or deblending would refute the claim that the accretion-heated bubble is observable.","tokens_in":23114,"feed_emoji":"🪐","tokens_out":9762,"duration_ms":102425,"temperature":0.7,"pith_summary":"Cold outer disk regions freeze CO out of the gas phase, leaving a midplane void between the two dragonfly-wing surfaces seen in CO channel maps. This paper argues that a luminous, actively accreting protoplanet inside that void heats its surroundings enough to sustain a roughly Hill-sphere-sized bubble of gas-phase CO, locally boosting the CO abundance by five orders of magnitude. In synthetic ALMA observations of a Jupiter-mass planet at 120 au, the bubble appears as a faint optically thick spot between the dragonfly wings, and after the rotating disk emission is subtracted it stands out as a 9-sigma point-source residual. If real disks behave this way, CO channel maps would give a largely automatic way to find forming planets even when no dusty circumplanetary disk is visible in the millimeter continuum. The same model explains why such continuum detections are rare: the horseshoe flow steadily drains large dust grains from the planet's vicinity.","feed_headline":"A warm CO bubble could reveal hidden protoplanets in ALMA data","feed_subtitle":"A 9-sigma residual after subtracting the disk flow marks the planet in synthetic C18O maps.","key_machinery":"The load-bearing object is the warm CO bubble, whose size is set by the balance between the planet's accretion luminosity $L = G M_p^2/(R_p t_{\\rm acc})$ and radiative cooling through two-population dust opacities: small submicron grains set the irradiation surface, while large millimeter grains regulate cooling. The bubble becomes observable through a freeze-out scaling: wherever $T \\le 21$ K the CO abundance is multiplied by $f_{\\rm freeze} = 10^{-5}$, so the $T > 21$ K region stands out against the frozen midplane. The detection pipeline is the subtraction of an axially symmetric kinematic model of the disk from the data cube, followed by median filtering, which turns the bubble into a point-source residual between the dragonfly wings.","core_discovery":"In the nominal 3D radiation-hydrodynamic run, a Jupiter-mass planet at 120 au outside the CO snowline, with accretion luminosity $L_p = 2.8 \\times 10^{-3}\\,L_\\odot$ (mass-doubling time 0.1 Myr and effective temperature about 4000 K), maintains a warm bubble in which the temperature stays above the 21 K freeze-out threshold out to slightly beyond the Hill radius. Because the model resets the CO abundance by a factor of $1/f_{\\rm freeze} = 10^5$ wherever $T > 21$ K, the bubble contains far more gas-phase CO than the surrounding frozen interior. Monte Carlo radiative-transfer post-processing of the same grid predicts that in C18O(2-1) channel maps this bubble is a low-intensity optically thick spot sitting in the void between the dragonfly wings; its emission intensity is nearly independent of isotopolog because unity optical depth is reached at the bubble's cool outskirts. After degrading to a 70 mas beam, adding ALMA-style thermal noise, and subtracting an axially symmetric kinematic disk model, the peak point-source residual reaches 9$\\sigma$ at +1.6 km/s (6$\\sigma$ at +1.2 and +1.4 km/s), so the planet would be detectable without a continuum counterpart. The same model shows the horseshoe flow removes millimeter-sized grains from the circumplanetary region, so the thermal continuum there is essentially absent, explaining the rarity of ALMA circumplanetary-disk detections.","pith_inferences":["Editorial inference: a bubble-size measurement would not translate directly into an accretion rate, because the bubble diameter also depends on the dust opacity law; surveys that use bubble size to rank accretion rates would need independent opacity constraints.","Editorial inference: the model implies young embedded giants should be continuum-bright early in their gap-clearing phase and then fade as horseshoe flows drain dust, so multi-epoch millimeter monitoring could distinguish newly formed planets from older ones.","Editorial inference: the quoted 9-sigma detection assumes roughly two days of integration under ideal calibration; realistic long-baseline observations with phase noise will likely need the median-filtered residual technique and favorable geometry, with inclination near 45 degrees and the planet off the disk minor axis, to reproduce the signal.","Editorial inference: if accretion luminosities near the assumed value are common during giant-planet formation, the same procedure could turn existing deep CO surveys of disks beyond roughly 70 au into automated protoplanet searches that do not wait for continuum or H-alpha detections."],"forward_implications":["In the nominal model the bubble is detectable in C18O(2-1) at 9$\\sigma$ peak after subtracting the axisymmetric disk flow, so CO channel maps can reveal an embedded giant planet even where the millimeter continuum and velocity kinks show nothing.","Because the bubble is optically thick, its brightness is almost independent of isotopolog; any CO isotopolog that is optically thin or weakly stratified across the gap, especially C18O, can serve as the tracer.","The model reproduces the observed scarcity of ALMA continuum detections of circumplanetary disks: horseshoe-flow depletion removes large dust within roughly 450 orbits, so the circumplanetary continuum fades with time.","Weaker accretion changes the picture: with a five-times-longer mass-doubling time the bubble shrinks to 45% of its nominal size, and a 20-Earth-mass planet with the same mass-doubling time yields a 30%-size bubble that is not detectable even in roughly 7-day integrations.","The bubble's size also depends on the dust opacity law; using uniform opacities recovers 73% of the nominal diameter for the same reduced luminosity, so the same observable can be produced by different combinations of luminosity and opacity."],"supporting_citations":[{"why":"Provides the accretion-luminosity heating formalism that sustains the warm bubble around the planet.","marker":"Benítez-Llambay et al. 2015"},{"why":"Establishes the dragonfly-wing channel-map geometry and the CO-freeze-out void in which the bubble must appear.","marker":"Dullemond et al. 2020"},{"why":"Supplies the 21 K freeze-out threshold used to scale the CO abundance in the radiative-transfer step.","marker":"Schwarz et al. 2016"},{"why":"Earlier work linking the accretion luminosity of embedded planets to observable chemistry, the route this paper extends to CO bubbles.","marker":"Cleeves et al. 2015"},{"why":"A recent model of accretion-heated circumplanetary chemistry that motivates the bubble scenario.","marker":"Jiang et al. 2023"},{"why":"Shows how accretion feedback modifies circumplanetary kinematics, compared here with the absence of detectable kinks.","marker":"Muley et al. 2024"},{"why":"Provides the isotopic ratios [12C]/[13C] = 77 and [16O]/[18O] = 560 used to build the 13CO and C18O synthetic images.","marker":"Wilson & Rood 1994"},{"why":"A steady-state analytical model predicting bright circumplanetary-disk continuum, the prediction the paper's dust-depletion result contradicts.","marker":"Zhu et al. 2018"}],"fun_headline_variants":["Warm CO bubble could reveal protoplanets in ALMA maps","ALMA 9-sigma CO signal traces hidden protoplanet","Protoplanet's warm CO bubble stands out in ALMA data","CO bubble around protoplanet predicted to glow in ALMA","Hidden protoplanets may shine via warm CO bubbles"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire bubble rests on how much energy the planet actually releases while swallowing gas: the luminosity is put in by hand as a 0.1 Myr mass-doubling time, not computed self-consistently, and the paper's own sensitivity test shows that a five-times-weaker release shrinks the bubble to 45% of its nominal size.","fun_headline_variants_meta":{"raw":{"variants":["Warm CO bubble could reveal protoplanets in ALMA maps","ALMA 9-sigma CO signal traces hidden protoplanet","Protoplanet's warm CO bubble stands out in ALMA data","CO bubble around protoplanet predicted to glow in ALMA","Hidden protoplanets may shine via warm CO bubbles"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000279,"raw_usage":{"total_tokens":1789,"prompt_tokens":1206,"completion_tokens":583,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":822,"completion_tokens_details":{"reasoning_tokens":495}},"tokens_in":822,"tokens_out":583,"duration_ms":7641,"temperature":1.0,"reasoning_tokens":495,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:13:16.479139+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for the bubble in a real disk with an independently confirmed accreting giant planet near 100 au: a C18O(2-1) observation at about 70 mas resolution, 0.2 km/s channels, and a deep integration should show a point-source residual above roughly 6$\\sigma$ at the planet's position after subtraction of the axisymmetric disk flow. A clean non-detection that cannot be blamed on viewing geometry or deblending would refute the claim that the accretion-heated bubble is observable.","supporting_citations":[{"cited_title":"P., Isella , A., Andrews , S","cited_arxiv_id":null,"evidence_quote":"Establishes the dragonfly-wing channel-map geometry and the CO-freeze-out void in which the bubble must appear."},{"cited_title":"R., Bergin , E","cited_arxiv_id":null,"evidence_quote":"Supplies the 21 K freeze-out threshold used to scale the CO abundance in the radiative-transfer step."},{"cited_title":"I., Bergin , E","cited_arxiv_id":null,"evidence_quote":"Earlier work linking the accretion luminosity of embedded planets to observable chemistry, the route this paper extends to CO bubbles."},{"cited_title":"D., & Klahr , H","cited_arxiv_id":null,"evidence_quote":"Shows how accretion feedback modifies circumplanetary kinematics, compared here with the absence of detectable kinks."},{"cited_title":"M., & Isella , A","cited_arxiv_id":null,"evidence_quote":"A steady-state analytical model predicting bright circumplanetary-disk continuum, the prediction the paper's dust-depletion result contradicts."}],"review_version":1}