{"id":"e1ab273c-63df-4eba-97c1-da0aadaa5078","arxiv_id":"2608.06181","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A two-way coupled model links bubble growth and resorption to magma flow, cooling, and friction, and argues the melt film thickness, not bubble radius, sets the growth regime.","lead":"This paper couples bubble-scale water diffusion with large-scale magma flow and cooling in a single numerical model, allowing bubbles to grow or shrink with the local conditions. The framework is meant to help volcanologists connect pyroclast textures to conduit and eruption dynamics.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The melt-film-thickness diffusion length scale underlying Eq. (3.1) is not implied by the spherical diffusion equation; the effective length is A S/(S-A), so the claimed regime-boundary shift needs an independent analytical or numerical check.","rationale":"I read the paper in good faith: it builds a modular two-way coupled framework (MVFFIN) that couples bubble-scale water diffusion with suspension-scale flow, and it demonstrates closed-system isothermal growth, suspension-viscosity effects, conduit-wall friction, and cooling-rind resorption. These are useful contributions, and the code is publicly available. The central claim, however, is the identification of the melt film thickness as the diffusional length scale that controls regime transitions. This claim is supported only by an assertion in Section 3.1 and by the authors' own simulations; it is not derived from Eq. (2.1) and is not tested against experimental data. A quasi-steady analysis of the spherical shell diffusion equation gives an effective length A S/(S-A), not S-A, so in the S>>A regime (which holds initially in Figs 6-7) the proposed Peb using (S-A)^2/D underestimates the diffusion time and can misplace the regime boundary. This is an internal-consistency issue, not a matter of disagreement with consensus. The monodisperse assumption flagged by the reader is real but is an acknowledged limitation; the length-scale issue is more fundamental because it threatens the qualitative claim itself. My proposed check settles it. If the check confirms the S-A scaling, the paper's central claim stands and the contribution is strengthened; if not, the regime diagrams and the headline conclusion need revision. Therefore the appropriate verdict is a conditional acceptance, with the diffusion-length check as a required condition.","tokens_in":24128,"tokens_out":18826,"duration_ms":222341,"concrete_test":"Solve the spherical diffusion equation (2.1) alone for a closed shell with no-flux at a=S and fixed equilibrium concentration at a=A, using Fig. 6 parameters (A0=3 mu m, S0=134 mu m, D=10^-12 m^2/s). Record the time tau_dep at which the cumulative diffusive flux into the bubble depletes 1/e of the shell's excess water, and compare tau_dep with (S-A)^2/D, A^2/D, and S^3/(D A). If tau_dep does not scale as (S-A)^2/D in the S>>A regime, then Eq. (3.1) mischaracterizes the diffusion timescale and the regime boundaries in Figs 7-9 would shift. Optionally, rerun the full coupled model with the same initial A0,S0 but with A0 varied relative to S0; if the transition time tracks A(S-A)/D rather than (S-A)^2/D, the central claim is falsified.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim (Section 6) is that the diffusion length scale for bubble growth is the melt film thickness (S-A), not the bubble radius A, 'substantially changing' regime-boundary predictions. This rests on the timescale tau_diff ~ (S-A)^2/D asserted in Section 3.1 without derivation ('structure of the diffusion equation and numerical experimentation'). The spherical diffusion equation (2.1) does not imply a film-thickness length scale. In a quasi-steady spherical shell with boundary concentrations c(A), c(S), the solution is c(r)=C1+C2/r, giving a gradient at the bubble wall grad(c)|_A = (c_S-c_A) S/[A(S-A)]. The diffusive flux scales as 4*pi*D*A*S*(c_S-c_A)/(S-A), so the effective diffusion length is A(S-A)/S, which reduces to A for S>>A and to S-A only when the film is thin (S approx A). For the simulations in Figs 6-7, S0 approx 134 mu m and A0 approx 3 mu m, so S>>A through most of the growth history; using (S-A)^2/D underestimates the diffusive supply time and can place the regime transition at the wrong time. The authors' numerical evidence is self-referential: the transition is identified from the same model, and the coincidence with Peb=1 (using S-A) is an observed correlation, not a test. Because the headline claim about where regime transitions occur is the main novelty, the choice of diffusion length scale is load-bearing and must be checked independently.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents MVFFIN, a modular multiscale numerical model that couples a single-bubble melt-shell model for volatile diffusion and bubble growth with one-dimensional suspension-scale mass, momentum, and energy equations in spherical (pyroclast) and cylindrical (conduit) geometries. The model incorporates water diffusion, outgassing by surface diffusion and permeable flow, temperature-dependent rheology, and fracture criteria. The authors use scaling analysis to define dimensionless regime markers and then simulate isothermal closed-system vesiculation, suspension-viscosity-limited growth, conduit-wall friction, and cooling-rind development. The central claims are that the melt film thickness, rather than the bubble radius, is the appropriate diffusion length scale for bubble-growth regime transitions; that suspension viscosity from crystals must be included in bubble-growth calculations; and that the coupled model reproduces cooling-rind resorption.","tokens_in":24421,"tokens_out":5845,"duration_ms":68033,"significance":"If the central claims hold, this is a potentially significant contribution to volcanological fluid dynamics: it provides an extensible, two-way-coupled framework linking bubble-scale degassing to suspension-scale flow, with an open-source implementation and a systematic dimensionless analysis. The use of literature constitutive laws without fitted parameters is a strength, as is the explicit treatment of thermal and mechanical feedbacks through a cooling rind. However, the headline claim about the melt-film-thickness diffusion length scale is not established by the equations as presented, the open-system regimes announced in the title and abstract are not simulated, and the quantitative stress/resorption results lack convergence and sensitivity tests. The framework is promising, but the broadest conclusions are not yet supported by the evidence in the manuscript.","major_comments":[{"comment":"The definition tau_diff ~ (S-A)^2/D is asserted on the basis of \"the structure of the diffusion equation and numerical experimentation,\" but it is not implied by the spherical diffusion equation (2.1). For a quasi-steady spherical shell, the steady solution of Eq. (2.1) is c(r) = C1 + C2/r, so the flux at the bubble wall scales as 4*pi*D*A*S*(c_S - c_A)/(S-A); the effective diffusion length is A(S-A)/S, which reduces to A for S >> A and to S-A only for S approximately A. In the simulations of Figs 6-7, S0 ~ 134 micrometers and A0 = 3 micrometers, so S >> A through most of the growth history. Using (S-A)^2/D can therefore underestimate the diffusion timescale by a large factor and shift the predicted Peb = 1 transition. Because the conclusion in Section 6 that the film thickness \"substantially changes the prediction for where regime transitions in bubble growth should occur\" is the paper's central novelty, this scaling must be checked by an independent analytical or high-resolution numerical solution of Eq. (2.1); the observed Peb = 1 coincidence in the same model is not a test. The regime diagrams in Figs 2-4 and the interpretation of Figs 6-7 should be revised on the basis of that check.","section":"Section 3.1, Eq. (3.1)"},{"comment":"The title and abstract present the paper as addressing \"open-system magmas\" and list outgassing through permeable porous networks and exposed magma-fluid interfaces among the regimes identified, and Section 2.4 does implement these processes. However, Section 6 states that numerical results documenting the effects of volatile loss through diffusive outgassing and gas percolation \"will be addressed in future works.\" All simulations in Section 4 are closed-system with respect to volatiles; Fig. 10 is thermally open but does not include volatile loss. The open-system regimes and the open-system portions of the regime diagrams in Fig. 3 are therefore not demonstrated. Please either provide the open-system simulations or explicitly narrow the title, abstract, and conclusions to closed-system and thermally open (cooling) behavior.","section":"Abstract, Section 6"},{"comment":"The quantitative claim that the hoop stress in the cooled rind reaches >4 GPa rests on the thin-walled elastic hoop-stress formula (2.27) and on the numerical viscosity cap of 10^12 Pa·s introduced in Section 2.2. The manuscript does not report the rind thickness h used in Eq. (2.27), does not provide a grid- or time-step-convergence study for the stress, and does not test sensitivity to the viscosity cap. Since the cap changes the effective solid-like response, the >4 GPa value and any fragmentation inference drawn from it are not robust as presented. Please add convergence tests for the adaptive time-stepping described in Section 2.3 and a sensitivity analysis to h and to the viscosity cap.","section":"Section 4.4, Eq. (2.27), Section 2.2"},{"comment":"The regime boundaries and resorption results are computed under the assumption that each grid node contains one isolated, radially symmetric spherical bubble with no nucleation, coalescence, Ostwald ripening, or crystal-bubble interactions, as stated in Sections 2.1 and 5.2. Section 6, however, generalizes the regime-transition predictions to \"magmatic systems\" without these restrictions. The limitations section is candid, but the central conclusions should be qualified to monodisperse, non-interacting bubble populations, or supplemented with a test of sensitivity to polydisperse populations or a simplified coalescence model, before the regime diagram is presented as a general predictive tool for high-vesicularity magmas where coalescence is known to be important.","section":"Sections 2.1, 5.2, 6"}],"minor_comments":[{"comment":"The text refers to an \"imposed relative velocity of 0.1\" and an \"artificial relative velocity,\" which appears to be a typo for \"relative viscosity\"; Section 4.1 uses the correct term.","section":"Section 4.2, Fig. 7"},{"comment":"The table lists η_r with units Pa·s, but η_r is dimensionless in Eq. (2.8a); the units should be corrected to \"1\" or \"dimensionless\" for consistency with η_infinity and η_0.","section":"Table 1"},{"comment":"The phrase \"stopped short of0 the fully-coupled model\" contains a stray character; it should read \"stopped short of the fully-coupled model.\"","section":"Section 5.3"},{"comment":"There are stray punctuation artifacts around Eq. (2.26), including a trailing period after the equation and a misplaced period in the equation numbering; these should be cleaned up.","section":"Section 2.5"},{"comment":"The notation Peb,A2 is introduced without definition; please define it explicitly or use a less ambiguous subscript.","section":"Section 4.1"},{"comment":"The reference list contains both \"Prousevitch et al. (1993)\" and \"Proussevitch et al. (1993)\" with different spellings of the first author; these should be unified and deduplicated.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is built on a broad and potentially valuable numerical framework, and the open-source code is a positive feature. The main risk is that the film-thickness scaling in Eq. (3.1) is presented as a finding rather than a hypothesis; if an independent check shows that the effective diffusion length is A(S-A)/S rather than (S-A), the regime diagrams and the central conclusion would change substantially. I would recommend asking for that check and for either open-system simulations or a clear narrowing of the title and abstract claims before further consideration."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read this with interest and think it is a real advance, not just a repackaging. The novelty is the two-way coupling of the Coumans bubble solver with suspension-scale flow, cooling, and resorption, plus the explicit treatment of suspension viscosity and conduit friction in the growth problem. The MVFFIN framework is modular, the code is on GitHub, and the regime diagrams in Figures 2–4 are useful organizing tools. The comparison of the coupled model to the decoupled bubble-scale limit is exactly the kind of sanity check that should be standard but often is not.\n\nThat said, the headline claim about the melt film thickness as the diffusion length scale does not sit right. The paper asserts in Section 3.1 that the structure of the diffusion equation and numerical experimentation give tau_diff ~ (S-A)^2/D, and then builds the bubble Peclet number Peb on that. But the steady-state spherical shell solution with fixed concentrations at A and S gives a bubble-wall flux scaling as A S/(S-A), so the effective diffusion length is A(S-A)/S, which reduces to A for thick films and only to S-A when the film is thin. For the simulations in Figs 6–7, with S0 ~ 134 microns and A0 ~ 3 microns, the system sits closer to the A-dominated end for much of the growth history. I am not claiming the paper is wrong, but the scaling is load-bearing for the regime-transition story, and it is presented without derivation. The numerical evidence is self-referential: the transition is identified from the same model that was used to infer the scaling. A referee should ask for an analytical derivation, or at least a systematic comparison of candidate length scales against the full numerical solution.\n\nThe rest of the soft spots are more standard. The title and abstract promise open-system behavior (diffusive outgassing, permeable flow), but the Results contain no simulations of those mechanisms; Section 6 explicitly defers them. There is no experimental validation and no convergence study. The cooling-rind hoop stress exceeding 4 GPa depends on a viscosity cap at 10^12 Pa·s that is not interrogated; that number should be shown to be insensitive to the cap. The monodisperse, non-interacting bubble assumption is acknowledged in Section 5.2, so I do not count it as a hidden flaw, but it does mean the regime boundaries are computed for an idealized texture.\n\nDespite these issues, the paper deserves a serious referee. The coupled model is a useful contribution and the regime framework will get used. My recommendation: send it out, but with the explicit instruction that the diffusion length-scale claim must either be derived or tested against independent numerical experiments, and the open-system mechanisms should be either demonstrated or clearly labeled as future work in the abstract and title.","headline":"A genuinely new two-way bubble-flow coupling with public code, but the headline film-thickness diffusion scaling needs a rigorous derivation or an independent numerical check before the regime boundaries are trusted.","tokens_in":24969,"tokens_out":3232,"would_cite":true,"duration_ms":41805,"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 argues that the melt film thickness between bubbles, not the bubble radius, sets the pace of diffusive bubble growth in magma, and that crystal-stiffened suspensions can limit vesiculation.","keywords":["bubble growth","magma degassing","vesiculation","melt film thickness","suspension viscosity","multiscale modeling","bubble resorption","Péclet number"],"falsifier":"A controlled decompression experiment with melts of identical composition, temperature, and starting bubble radius but different bubble number densities, hence different melt film thicknesses, would settle the claim: if the onset of diffusion-limited growth tracks bubble radius rather than film thickness, the rescaling fails, while collapse of the growth-rate transition onto a curve in $(S-A)^2/D$ would support it.","tokens_in":1737,"feed_emoji":"🌋","tokens_out":1774,"duration_ms":84301,"temperature":0.7,"pith_summary":"The paper builds a model that couples growth of individual water-vapor bubbles in magma to the larger-scale flow, cooling, and outgassing of the magma body, so that bubble expansion and collapse emerge from local conditions. Its central claim is that the length scale controlling how fast water diffuses into a bubble is the thickness of the melt film between bubbles, not the bubble radius, and that this rescaling moves the predicted boundary between viscous-limited and diffusion-limited bubble growth. It also argues that crystals stiffen the suspension and can make the suspension itself the bottleneck for bubble growth. If these claims hold, predictions of where magma vesiculates, overpressurizes, and fragments should be revised, and laboratory vesiculation experiments can be interpreted with a single model.","feed_headline":"Diffusive bubble growth scales with melt film thickness, not radius","feed_subtitle":"If right, predicted regime transitions for degassing magma shift, and crystals can choke bubble growth.","key_machinery":"The central object is the multiscale shell model: each computational node carries one representative spherical bubble of radius $A$ inside a melt film of radius $S$, with water diffusion solved in the film and bubble radius advanced by viscous resistance, while the film thickness $S-A$ is set by bubble number density. The key identity is the rescaled bubble Péclet number $\\mathrm{Pe}_b = \\Delta P_b (S-A)^2/(\\mu D)$, which replaces the classic radius-based form. The numerical coupling passes vesicularity changes into the suspension-scale compressible-flow and energy equations as volumetric source terms, and returns updated pressure, temperature, and water content to the bubble solver at every time step.","core_discovery":"The paper's central claim is that in a bubbly magma the relevant diffusive length for volatile exchange is the melt film thickness $S-A$, not the bubble radius $A$, so the bubble Péclet number should be $\\mathrm{Pe}_b = \\Delta P_b (S-A)^2/(\\mu D)$. Because $S-A$ is set by bubble number density, regime transitions in bubble growth depend on number density more directly than on bubble size. Second, bubble growth is resisted not only by melt viscosity but by the effective suspension viscosity, including the contribution of rigid crystals. Third, coupling bubble-scale diffusion to suspension-scale momentum, heat, and volatile transport allows cooling rinds to suppress vesiculation and eventually drive resorption of the vesicles. The model is implemented in the modular MVFFIN numerical framework and, in the paper's view, provides a basis for interpreting laboratory experiments and conduit-scale volcanic processes.","pith_inferences":["Beyond the paper: if the melt film thickness is the controlling diffusion length, a polydisperse bubble population cannot be represented by a single regime transition; bubbles with the same radius but different local film thicknesses should cross from viscous-limited to diffusion-limited growth at different times.","Beyond the paper: the suspension-viscosity mechanism implies that decompression experiments on crystal-free melts will systematically under-predict vesiculation lag in crystal-bearing magmas; this could be tested by repeating decompression experiments with increasing crystal content.","Beyond the paper: the cooling-rind mechanism suggests an engineered-foam analog test, since a thin, stiff, poorly permeable crust should suppress interior bubble expansion and eventually reverse it in bread or polyurethane systems as well.","Beyond the paper: the rescaled Péclet number implies that the transient early-growth regime is governed by a $\\sqrt{Dt}$ diffusion length, so near-instantaneous solubility changes matter mainly for the earliest stage of bubble response, not for the long-time regime boundary."],"forward_implications":["Regime boundaries for bubble growth in magmatic systems should be evaluated with the melt film thickness in the bubble Péclet number, not the bubble radius, which shifts where transitions are predicted.","Bubble growth in crystal-bearing magmas should include suspension viscosity, so crystals can impede vesiculation even when they are large compared with the bubbles.","The coupled model reproduces cooling-rind suppression of vesiculation and predicts that continued cooling can force full resorption of vesicles in pyroclast margins.","The regime analysis separates controls from bubble-scale diffusion, viscous melt resistance, suspension transport, outgassing, and thermal quenching, giving a map of which process dominates in a given magma geometry.","The modular code is extensible to conduit flow, pyroclast evolution, laboratory experiments, and other vesiculating materials such as foams and doughs."],"supporting_citations":[{"why":"supplies the experimentally validated bubble-growth solver that this model adapts for repeated restarting and coupling.","marker":"Coumans et al. (2020)"},{"why":"established the two limiting regimes of bubble growth, viscous versus diffusive, that the paper rescales.","marker":"Navon et al. (1998)"},{"why":"provides the classic isothermal diffusive bubble-growth dynamics and the transient $\\sqrt{Dt}$ diffusion length.","marker":"Prousevitch et al. (1993)"},{"why":"documents radial melt-viscosity variations around growing bubbles and gas overpressure in the melt shell.","marker":"Lensky et al. (2001)"},{"why":"shows how cooling margins quench vesicular magma fragments, the physical basis for the rind mechanism.","marker":"Kaminski and Jaupart (1997)"},{"why":"supplies the two-phase magma rheology and capillary-number treatment used in the suspension viscosity.","marker":"Mader et al. (2013)"},{"why":"provides the melt viscosity model used in the example simulations.","marker":"Hess and Dingwell (1996)"},{"why":"provides the water diffusivity model used in the bubble-scale and outgassing computations.","marker":"Zhang and Ni (2010)"},{"why":"defines the rhyolite composition and breadcrust-bomb context used for the demonstration simulations.","marker":"Wright et al. (2007)"},{"why":"provides the permeability relationship used for permeable gas flow in the open-system treatment.","marker":"Mueller et al. (2005)"}],"fun_headline_variants":["Bubble growth scales with melt film thickness, not radius","Cooling rinds reverse bubble growth, driving resorption","Suspension viscosity and crystals resist magma bubble growth","Bubble dynamics shift with number density, not size alone","Melt film thickness governs bubble Péclet number in magma"],"cache_read_input_tokens":27008,"weakest_assumption_plain":"Every computational point is treated as one isolated spherical bubble inside a shell of melt, with no new bubbles forming, no bubbles merging, and no crystals touching the bubble; if those omissions change how water reaches bubbles or how the suspension resists flow, the predicted regime boundaries and resorption timing shift.","fun_headline_variants_meta":{"raw":{"variants":["Bubble growth scales with melt film thickness, not radius","Cooling rinds reverse bubble growth, driving resorption","Suspension viscosity and crystals resist magma bubble growth","Bubble dynamics shift with number density, not size alone","Melt film thickness governs bubble Péclet number in magma"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000222,"raw_usage":{"total_tokens":1490,"prompt_tokens":1015,"completion_tokens":475,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":631,"completion_tokens_details":{"reasoning_tokens":394}},"tokens_in":631,"tokens_out":475,"duration_ms":5161,"temperature":1.0,"reasoning_tokens":394,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:04:55.138790+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A controlled decompression experiment with melts of identical composition, temperature, and starting bubble radius but different bubble number densities, hence different melt film thicknesses, would settle the claim: if the onset of diffusion-limited growth tracks bubble radius rather than film thickness, the rescaling fails, while collapse of the growth-rate transition onto a curve in $(S-A)^2/D$ would support it.","supporting_citations":[{"cited_title":"1998 , journal =","cited_arxiv_id":null,"evidence_quote":"established the two limiting regimes of bubble growth, viscous versus diffusive, that the paper rescales."},{"cited_title":"1997 , journal =","cited_arxiv_id":null,"evidence_quote":"shows how cooling margins quench vesicular magma fragments, the physical basis for the rind mechanism."},{"cited_title":"2013 , journal =","cited_arxiv_id":null,"evidence_quote":"supplies the two-phase magma rheology and capillary-number treatment used in the suspension viscosity."},{"cited_title":"2010 , journal =","cited_arxiv_id":null,"evidence_quote":"provides the water diffusivity model used in the bubble-scale and outgassing computations."},{"cited_title":"and Cashman, Katharine V","cited_arxiv_id":null,"evidence_quote":"defines the rhyolite composition and breadcrust-bomb context used for the demonstration simulations."},{"cited_title":", number =","cited_arxiv_id":null,"evidence_quote":"provides the permeability relationship used for permeable gas flow in the open-system treatment."}],"review_version":1}