{"id":"4c101ee5-6f74-44b0-8426-96baada34cb0","arxiv_id":"2505.00363","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Self-gravitating spiral waves can concentrate a realistic multi-size dust population enough to collapse into bound planetary cores when the dust-to-gas ratio is about 0.01 and grains reach Stokes numbers near unity.","lead":"This paper uses 3D shearing-box simulations to show that spiral density waves in a gravitationally unstable protoplanetary disc can concentrate dust by more than an order of magnitude, even when the dust spans a wide range of particle sizes. If the dust-to-gas ratio is near 1 percent and some grains have grown to roughly centimeter sizes, the concentrated dust can collapse directly into planetary cores of 1 to 10 Earth masses.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'bound clumps' in the abstract are identified only via a Roche surface-density threshold (Eq. 14, Sec. 3.2.3); no virial or binding-energy check is reported, so the direct-collapse step of the central claim is not yet demonstrated.","rationale":"The paper is a careful numerical parameter study and the multi-size extension is a genuine step beyond single-Stokes-number work. I do not dispute the measured density enhancements: the test-particle run and the massive-particle runs show order-of-magnitude and larger surface-density contrasts, and the authors are appropriately cautious in their concluding caveat about cm-sized growth during the short self-gravitating phase. That caveat is real but it concerns applicability to real discs; it does not test whether the simulated mechanism works. My concern targets the stronger assertion, in the abstract and conclusion, that these overdensities 'undergo direct gravitational collapse to form bound clumps'. The clump-finding procedure in Sec. 3.2.3 is a Roche-density threshold, not a bound-state measurement. Since the run already includes particle self-gravity, evaluating the virial parameter of each reported clump is a cheap and decisive check. If the clumps are bound and contracting, the central claim stands and the CONDITIONAL verdict would rest only on the grain-growth caveat; if not, the paper should be revised to claim strong concentration and possible gravitational collapse, rather than demonstrated core formation. The reader's rationale mentions the same threshold issue as a secondary point, so we partially agree; I make it the primary concern because it bears on whether the claimed phenomenon occurs at all, before any question of grain growth.","tokens_in":19188,"tokens_out":9274,"duration_ms":105986,"concrete_test":"Using the publicly archived snapshots from a clump-producing run (e.g. md01St002-200 at tOmega=300), compute for every super-Roche region the actual virial parameter 2T/|W| including the tidal term and the gas-drag contribution, and follow the region for 10 Omega^{-1}; require 2T/|W| < 1 and continued contraction for the region to count as a bound clump. This single check distinguishes genuine gravitational collapse from transient density enhancements.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that spirals directly produce bound planetary cores depends on identifying particle overdensities as gravitationally bound clumps. In Sec. 3.2.3 clumps are selected with a surface-density threshold: a region is called a clump when its projected particle surface density exceeds Sigma_R = 8.8 Omega^2 H_d/G, approximated as 10, with H_d ~ 1 assumed. This is a local tidal-stability criterion, but it is not a demonstration that the region has actually collapsed. A transient aerodynamic concentration in a spiral arm can exceed this threshold while remaining unbound in three dimensions, still being sheared apart, or being supported by gas pressure and diffusion for the smaller grains. The clump masses (0.6-7.15 M_Earth) are then computed by summing superparticles inside the Hill radius around each density peak, so the mass scale is partly set by the threshold definition rather than by a dynamically verified bound state. No convergence study is shown, so it is unknown whether these super-Roche regions survive as bound objects at higher particle number or grid resolution. This matters because the abstract's 'bound clumps' and the conclusion's 'gravitationally collapse to directly form dense clumps' are the strongest form of the central claim. The simulations robustly show strong dust concentration; the collapse step itself is inferred from a threshold, not verified.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents 3D shearing-box simulations with the Pencil Code of a dust population in a self-gravitating protostellar disc, using a multi-size particle distribution n(a)∝a^-4 with representative Stokes numbers spanning about 0.02–200, and including particle self-gravity and back-reaction. The test-particle run shows that spiral density waves enhance the local dust surface density by more than an order of magnitude, with the strongest concentration for particles near St~1. In runs with total dust-to-gas ratio of order 0.01, or about 0.003 for narrow Stokes-number ranges centred near unity, the simulations produce regions where the projected particle surface density exceeds about 10 times the mean gas surface density; 42 such regions are identified as clumps and, under a scaling to a 1 M_sun star at 50 AU, have masses between 0.6 and 7.15 M_earth. The authors conclude that self-gravitating spirals can directly form bound planetary cores, bypassing the metre barrier, while acknowledging in Section 5 that it is not clear whether grain growth can reach cm sizes within the short (~1e5 yr) self-gravitating phase.","tokens_in":19452,"tokens_out":9975,"duration_ms":100410,"significance":"If the central claim is established, the paper offers a credible route to forming ~1–10 M_earth cores in the first ~1e5 yr of disc evolution, potentially explaining early planet formation and bypassing the drift barrier. The study's strengths are the multi-size dust treatment, the systematic parameter sweep over dust-to-gas ratio and Stokes-number range, and the inclusion of particle self-gravity and back-reaction; the authors also state that data and scripts will be archived and that the Pencil Code is public. The significance is, however, conditional: the direct-collapse conclusion rests on a projected surface-density threshold rather than on a dynamical demonstration of boundness, no convergence study is reported, and the applicability to real discs depends on cm-sized grains being present during the brief self-gravitating phase, a caveat the paper states honestly.","major_comments":[{"comment":"The only clump-identification criterion is a projected particle surface density exceeding Σ_R≈10⟨Σ_g⟩. This is a Roche/tidal threshold, not a demonstration that the region has undergone gravitational collapse and is bound. A transient aerodynamic concentration in a spiral arm can satisfy this projected-density criterion while remaining unbound in three dimensions, supported by gas drag or diffusion, or subject to tidal shear. Because the simulations include particle self-gravity and gas gravity, this can be tested directly: for each candidate, compute the particle self-binding energy or a virial-type ratio, compare the three-dimensional particle density with the Hill density of Eq. (12), and follow the clump evolution to see whether it contracts and survives. The abstract's 'bound clumps' and the conclusion's 'gravitationally collapse to directly form dense clumps' are stronger than what a surface-density threshold demonstrates, so this point is load-bearing for the central claim.","section":"Section 3.2.3, Eq. (14)"},{"comment":"The clump masses are obtained by summing superparticles within a Hill radius defined around each density peak, and the Hill radius is itself a function of the clump mass (R_Hill=(m_cl/3)^{1/3} in code units). The reported mass range 0.6–7.15 M_earth is therefore partly set by the detection algorithm rather than by a dynamically determined bound state. Please provide radial enclosed-mass profiles, the peak volume density relative to the Hill density, and a binding-energy estimate for representative clumps, so that the '1–10 M_earth planetary cores' claim is supported by clump structure rather than by the threshold geometry.","section":"Section 3.2.3, Fig. 10"},{"comment":"No resolution or particle-number convergence study is reported; all runs use a 256×256×128 grid and 10^7 superparticles, and clump detection uses a 1000×1000 surface-density grid. It is therefore not established that the super-Roche regions survive as bound clumps at higher resolution or with different particle sampling, and the clump masses could be resolution-dependent. Please add at least one clump-forming case at higher grid resolution and particle number, and show that the clump-formation criterion and the mass distribution are stable; ideally also test a different box size to check that the periodic shearing box does not artificially promote clumping.","section":"Sections 2 and 3"}],"minor_comments":[{"comment":"The conversion from Hill volume density to surface density uses Σ_Hill∼2H_dρ_Hill; for a Gaussian vertical profile the numerical factor is sqrt(2π)≈2.51, and the resulting value Σ_Hill≈1.4 should be derived explicitly for the assumed H_d=1.","section":"Section 3.2.2, Eq. (13)"},{"comment":"The threshold is rounded from Σ_R=8.8 to 10, and the text states that this corresponds to an enhancement of 8.8/0.01=880 times the mean dust surface density for m_d/m_g=0.01; with the adopted threshold of 10 the required enhancement is 1000. Please make the final threshold and its normalization consistent.","section":"Section 3.2.3"},{"comment":"The statement that 'if the Stokes number 1 particles have sizes of order a cm, then our broad size distribution extends from about 0.1 mm to about 1 m' does not match the stated Stokes range 0.02–200, which would extend to about 2 m for a 1 cm St=1 grain; please check the size conversion.","section":"Section 3.1.1"},{"comment":"The caption of Figure 9 appears truncated in the version I reviewed, with the bottom-panel sentence ending at 'mean gas surface density' and an uncompleted formula; please ensure the published caption is complete.","section":"Figure 9 caption"},{"comment":"The data availability statement contains a grammatical error ('will be archived will be publicly available') and should be corrected.","section":"Data availability statement"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful and well-conducted parameter study, and the multi-size dust treatment is a genuine step forward. My main concern is that the abstract and conclusions assert bound-clump formation and direct collapse, while the presented evidence is a projected surface-density threshold. I would be comfortable with acceptance after the authors add post-processing boundness checks, a convergence statement, and, if those checks cannot be performed within the current data, soften the wording of the central claim. I would not recommend rejection because the underlying simulations and trends are sound and the missing checks are achievable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a careful, useful extension of earlier single-size work. The main simulation result holds up; the weak spot is that 'bound clumps' are identified with a Roche surface-density threshold, not a binding-energy check, so the collapse step is inferred rather than demonstrated. It deserves serious refereeing, but the strongest claim needs either a better clump test or softer language.\n\nWhat's new: first (to my knowledge) shearing-box study of dust in self-gravitating spirals with a broad power-law size distribution (St~0.02-200), including dust self-gravity and back-reaction. The central result is robust: even with most mass in small grains, spirals still concentrate dust by over an order of magnitude, and dense clumps appear when dust-to-gas ratio ~0.01 and the distribution extends to St~1. The parameter sweep is well designed - test particles first, then massive particles, then narrower size ranges - and the clumping criterion is cleanly stated as a mass-in-near-unity-Stokes requirement. The citation pattern is appropriate, and the paper commits to archiving data and scripts.\n\nSoft spots: the clump finder uses the Roche surface density Sigma_R = 8.8 Omega^2 H_d/G with H_d~1, rounded to 10 times the mean gas surface density. That's a reasonable tidal-stability criterion, but it doesn't show that a region has actually collapsed. A transient concentration could exceed the threshold yet remain unbound or shear apart. Since the abstract and conclusions treat these as 'bound clumps' that 'directly form' cores, this is a real gap. Clump masses are also defined by summing superparticles inside the Hill radius, so the 1-10 M_Earth range is partly set by the selection rule. I'd want a virial/binding-energy check, or the language softened. No convergence study is shown either, so we don't know how clump statistics depend on resolution and particle number.\n\nThe other caveat is the authors' own: it's not clear grains can grow to cm sizes within the first ~1e5 years of the self-gravitating phase. They cite plausible evidence for early growth, but applicability remains conditional. That's correctly stated, not a flaw in the simulations.\n\nBottom line: for people working on dust dynamics and planet formation in young massive discs, this is a solid incremental step, not a paradigm shift. The soft spots are addressable. Send it to peer review - with a request for a clump binding check or softer language, and a resolution test if possible.","headline":"Solid multi-size extension of the GI-dust story, but the 'bound clumps' claim rests on a Roche threshold rather than a binding check, so the strongest conclusion is one step ahead of the evidence.","tokens_in":20021,"tokens_out":5369,"would_cite":true,"duration_ms":49419,"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 paper argues that self-gravitating spiral density waves in young protoplanetary discs can concentrate a broad size range of dust enough for the solid component to collapse directly into bound planetary cores of 1 to 10 Earth masses.","keywords":["protoplanetary discs","gravitational instability","dust concentration","planetesimal formation","planetary core formation","metre barrier","shearing-box simulations","Stokes number"],"falsifier":"Submillimetre and radio surveys of the youngest Class 0/I protostellar discs at tens of AU that rule out millimetre- to centimetre-sized grains would directly test the premise, because without those large grains the spirals never receive the particles they need to concentrate.","tokens_in":18946,"feed_emoji":"🪐","tokens_out":7824,"duration_ms":78664,"temperature":0.7,"pith_summary":"Planet formation by core accretion struggles with the metre barrier: in a smooth disc, gas drag makes centimetre-to-metre particles spiral into the star before they can grow. This paper argues that a young disc's self-gravitating spirals break that bottleneck. Three-dimensional shearing-box simulations with a realistic range of dust sizes show the spirals raise local solid surface density by more than an order of magnitude, concentrating intermediate-size grains in narrow filaments. If the dust holds about one per cent of the gas mass and includes grains with Stokes numbers near unity, those concentrations cross the Roche threshold and collapse into bound clumps of roughly 1 to 10 Earth masses. The route would form planetary cores, or the cores of giant planets, within the first roughly 100,000 years of the disc's life, bypassing the slow collisional path.","feed_headline":"Spiral arms can collapse dust straight into 1-10 Earth-mass cores","feed_subtitle":"Simulations show young discs' spirals concentrate dust past the Roche threshold to form cores within the first ~100,000 years.","key_machinery":"The mechanism is the coupling between self-gravitating spiral density waves and aerodynamic drag. The Stokes number, $\\mathrm{St}=\\tau_s\\Omega$, measures how long a particle's drag stopping time $\\tau_s$ is relative to the orbital time; particles with $\\mathrm{St}\\sim 1$ feel gas drag strongly enough to drift toward the spiral's pressure maxima but not so strongly that they diffuse away. These particles also settle to a thin layer near the midplane, so their local surface density can be enhanced by two orders of magnitude. Collapse is diagnosed with the Hill/Roche surface density, $\\Sigma_R\\simeq 8.8\\,\\Omega^2 H_d/G$ (with $H_d$ the dust scaleheight), approximated as ten times the mean gas surface density; the simulated clump masses are then converted to Earth masses through a shearing-box scaling to a 50 AU disc around a solar-mass star.","core_discovery":"The paper's central claim is that gravitational collapse of the solid component can occur inside the spiral density waves of a self-gravitating disc even when the dust has a broad size distribution. In the simulations, the dust follows a power-law size distribution $n(a)\\propto a^{-4}$ spanning four orders of magnitude in particle size (representative Stokes numbers $0.02$ to $200$), yet the spirals still increase the local particle surface density by more than an order of magnitude. Particles with $\\mathrm{St}\\sim 1$ settle into a thin midplane layer and drift toward pressure maxima, producing the strongest enhancements. When the dust-to-gas ratio is $\\sim 0.01$ for the full size range, or when particles with Stokes numbers between roughly $0.5$ and $5$ carry a few times $10^{-3}$ of the gas mass, local surface densities exceed the Roche value and collapse into bound clumps. Across the runs that formed clumps, 42 clumps emerged with masses from $0.6$ to $7.15\\,M_\\oplus$ (mean $2.2\\,M_\\oplus$), in line with the 1–10 Earth-mass cores predicted by two-fluid disc studies.","pith_inferences":["A testable extension of this result is that if direct collapse is the main core-forming route, the observed super-Earth mass distribution should peak near a few Earth masses, echoing the simulated clump-mass mean of about $2\\,M_\\oplus$.","The concentration mechanism may bootstrap grain growth: even small grains are gathered by the spirals, so a disc that starts with only micron dust could raise its own largest grain sizes into the $\\mathrm{St}\\sim 1$ regime during the self-gravitating epoch.","The same pressure-maximum concentration logic might operate in other gravitoturbulent environments, but the required particle sizes and cooling times there would have to be evaluated independently.","Because the simulations use a shearing box, whether the collapsed clumps survive tidal shear when followed through a full orbit in a global disc remains an open extension of this work."],"forward_implications":["If the mechanism operates, planetary cores of roughly $1{-}10\\,M_\\oplus$ can appear within the first $\\sim 10^5$ years, before the disc loses its self-gravitating state.","The metre barrier stops being fatal: grains need only reach centimetre sizes to be concentrated, after which direct collapse can form cores instead of a long collisional cascade.","Even when no clumps form, spiral-induced enhancements of more than an order of magnitude could accelerate grain growth by raising local dust densities and collision rates.","Formation of super-Earths and of the solid cores of giant planets becomes possible during the Class 0 phase, matching observations of accreting protoplanets in slightly older discs.","The required dust-to-gas ratio is close to the canonical value of $0.01$ for a broad size range, and only about $0.003$ for particles with Stokes numbers between $1$ and $10$."],"supporting_citations":[{"why":"Establishes the gas-drag drift that creates the metre barrier and supplies the Epstein drag law used for the particles.","marker":"Weidenschilling 1977"},{"why":"Shows that self-gravitating spirals can concentrate solid particles and motivates the drag-regime treatment.","marker":"Rice et al. 2004"},{"why":"Earlier single-size simulations showing density enhancement depends on particle stopping time and that dense clumps can form.","marker":"Gibbons et al. 2014"},{"why":"Shearing-box study of dust in self-gravitating discs whose setup, with $Q\\sim 1$ and $\\beta$-cooling, is followed here.","marker":"Baehr & Zhu 2021"},{"why":"Provides the Hill/Roche surface density criterion used to identify bound clumps and earlier evidence for clump formation.","marker":"Baehr et al. 2022"},{"why":"Two-fluid disc modelling that predicts solid collapse yields 1–10 Earth-mass cores.","marker":"Longarini et al. 2023b"},{"why":"Independent two-fluid simulations supporting the same 1–10 Earth-mass core outcome.","marker":"Rowther et al. 2024a"},{"why":"Supplies growth timescale estimates ($\\sim 1/(Z\\Omega)\\sim 10^4$ yr at 50 AU) used to argue grains can reach centimetre sizes during the self-gravitating phase.","marker":"Birnstiel et al. 2012"},{"why":"Provides the physical scaling from shearing-box units to a 50 AU solar-mass disc, converting simulated clump masses to Earth masses.","marker":"Schäfer et al. 2017"}],"fun_headline_variants":["Spiral waves collapse dust to build Earth-mass cores","Dust collapse in spirals bypasses growth barrier to form cores","Gravitational instabilities in discs directly form planetary cores","Spirals turn dust into 1-10 Earth-mass clumps","Direct core formation via dust collapse in self-gravitating discs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole route depends on dust grains growing to roughly centimetre sizes during the first 100,000 years of disc life, when the disc is still self-gravitating; the paper itself cautions that it is not clear grain growth can be that fast.","fun_headline_variants_meta":{"raw":{"variants":["Spiral waves collapse dust to build Earth-mass cores","Dust collapse in spirals bypasses growth barrier to form cores","Gravitational instabilities in discs directly form planetary cores","Spirals turn dust into 1-10 Earth-mass clumps","Direct core formation via dust collapse in self-gravitating discs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000631,"raw_usage":{"total_tokens":2957,"prompt_tokens":1034,"completion_tokens":1923,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":650,"completion_tokens_details":{"reasoning_tokens":1837}},"tokens_in":650,"tokens_out":1923,"duration_ms":14455,"temperature":1.0,"reasoning_tokens":1837,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:44:15.084087+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Submillimetre and radio surveys of the youngest Class 0/I protostellar discs at tens of AU that rule out millimetre- to centimetre-sized grains would directly test the premise, because without those large grains the spirals never receive the particles they need to concentrate.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Hill/Roche surface density criterion used to identify bound clumps and earlier evidence for clump formation."}],"review_version":1}