{"id":"0899ab71-702d-43e8-9631-2c6334d3261b","arxiv_id":"2504.16401","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the 3D Boussinesq system near Couette flow with constant background temperature, H2 perturbations satisfying velocity and temperature smallness bounds of order ν and ν² respectively stay global in time.","lead":"This paper proves that small disturbances to the Couette shear flow in a three-dimensional Boussinesq fluid remain globally regular when the velocity perturbation is at most of order viscosity and the temperature perturbation is at most of order viscosity squared. The result establishes a Sobolev-space stability threshold for the constant-temperature case, where the 3D lift-up mechanism is active.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Omitted proof of Prop. A.2 is the load-bearing risk: E4.1 and all subsequent closures depend on this imported coupled linear estimate, which is stated without derivation.","rationale":"Read in good faith, the paper is a serious quasi-linearization argument for the 3D Boussinesq threshold, and the main structure (decomposition of u1,0, good unknowns Q, auxiliary energy E7) is coherent. I checked the circularity that worried the reader: the smallness condition ∥du1,0∥H4+ν^{-1}∥∂tdu1,0∥H2<δ is indeed implied by the bootstrap assumption E1≤ε0, so using E1 while proving E1 is a standard bootstrap, not a logical circle. The absence of a local well-posedness statement is a presentation gap; the heuristic optimality discussion is not load-bearing. The genuinely load-bearing weakness is the unproven imported coupled linear estimate Proposition A.2, used at the entry point of the nonlinear closure. Because the paper's own text concedes the proof is omitted, and because the displayed application has a weight mismatch, this is the right place to demand a concrete check. If the re-derivation succeeds (or the mismatch is repaired), the conditional verdict can be upgraded; if it fails, the theorem is unproven. I therefore keep the reader's CONDITIONAL verdict unchanged.","tokens_in":45114,"tokens_out":17163,"duration_ms":155301,"concrete_test":"Independently re-derive Proposition A.2 from Proposition 4.3 of Wei-Zhang [35], writing out the full proof and verifying P0f=P0f1=P0f2=P0g=P0g1=0 for f=△u2,≠ and g=u3,≠ with the Boussinesq forcings. Then replace the RHS norm in (6.1) with the weights actually controlled by E5 (Xb for ∂x²Θ≠, Xa for ∂z²Θ≠) and confirm the resulting E4² bound is still absorbed under E4≤ε0ν, E5≤ε0ν², E6≤ε0ν. If either step fails, Theorem 1.1 is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is a bootstrap over E1-E6. The most exposed step is the estimate of E4.1 in Proposition 6.1, which applies Proposition A.2 to the coupled system for f=△u2,≠ and g=u3,≠. Proposition A.2 is not proved; the text in Appendix A says only 'can be derived from Proposition 4.3 in [35], and we omit it.' This is not a cosmetic gap: A.2 contains the nonlocal lift-up coupling -2∂x∂z△^{-2}f and a forcing (∂x²+∂z²)Θ≠, and it must hold with exactly the zero-mode conditions and initial-data terms used in (6.1). If A.2 fails, or if its derivation from [35] requires hypotheses not verified for the Boussinesq forcings, then the E4 bound is unsupported; since E4 enters E1 (Prop. 5.1), E2, E5 and E6, the bootstrap and Theorem 1.1 collapse. A secondary indicator that the imported estimate has not been fully re-checked is the mismatch in (6.1): the RHS uses ν^{-4/3}||(∂x²,∂z²)Θ≠||²_Xb, although E5 only gives ∂z²Θ≠ in Xa. This is likely fixable, but it shows the A.2 application needs a complete derivation before the closure can be certified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the 3D Boussinesq system near Couette flow in T×R×T with ν=μ and g=1. The main theorem (Theorem 1.1) asserts that if the initial velocity perturbation is H²-bounded by εν and the initial temperature perturbation by εν², for a universal ε>0, then the solution is global in time. The proof is organized as a bootstrap over six energy functionals E1–E6, combining zero-mode and non-zero-mode estimates. The non-zero-mode estimates rely on space-time linear estimates for the perturbed linear operator L_V that are imported from Wei–Zhang [35], with one new coupled estimate, Proposition A.2, stated without proof. The paper claims the exponents β=1 for velocity and β=2 for temperature are optimal, supported by the formal balance in Remark 1.2.","tokens_in":45392,"tokens_out":5004,"duration_ms":47551,"significance":"If the proof is completed, this would be the first Sobolev-space stability threshold for the unstratified 3D Boussinesq Couette problem, where the 3D lift-up effect is active, complementing the stratified result of Coti Zelati–Del Zotto–Widmayer [14]. The bootstrap architecture is transparent, the paper explicitly identifies the hard imported linear estimates, and the formal optimality discussion in Remark 1.2 gives a concrete falsifiable prediction. The introduction of the new good unknown Q=u_{2,≠}+κu_{3,≠} and the quasi-linear decomposition are plausible and potentially useful ingredients. The paper is also honest about its reliance on external results. However, as detailed below, a load-bearing gap appears in the treatment of the coupled linear estimate Proposition A.2, and several smaller technical issues must be addressed before the proof is certifiable.","major_comments":[{"comment":"Proposition A.2 is the key imported estimate for the coupling between △u_{2,≠} and (∂x²+∂z²)u_{3,≠}, and it is used in the derivation of the E4 bound in Proposition 6.1. The proposition is stated with the sentence \"can be derived from Proposition 4.3 in [35], and we omit it,\" with no proof or derivation. This is not a cosmetic omission: the proposition contains the nonlocal lift-up coupling term −2∂x∂z△^{-2}f and a temperature forcing, and it must hold with exactly the zero-mode conditions and initial-data terms used in (6.1). If the derivation from Proposition 4.3 of [35] requires hypotheses that are not verified for the Boussinesq forcings, then the E4 bound is unsupported; since E4 enters the closures for E1, E2, E5 and E6, Theorem 1.1 collapses. Please provide a complete proof of Proposition A.2, or at minimum a detailed derivation from Proposition 4.3 of [35] that verifies every hypothesis, including the smallness condition on du_{1,0} and the zero-mode conditions for all forcing terms.","section":"Appendix A, Proposition A.2; used in §6.1, Eq. (6.1)"},{"comment":"The right-hand side of (6.1) contains the term ν^{-4/3}∥(∂x²,∂z²)Θ_{≠}∥²_{X_b}, but the energy E5 defined in Section 2.2 only gives ∥∂x²Θ_{≠}∥_{X_b} and ∥∂z²Θ_{≠}∥_{X_a}, with 0<a<b. With the given definitions, the pair (∂x²Θ_{≠}, ∂z²Θ_{≠}) is not directly bounded in X_b by E5. This mismatch must be fixed, for instance by proving Proposition A.2 with the temperature forcing in the appropriate X_a norm, or by adding an additional term to E5. As written, the line after (6.1) \"using Lemma 4.3, Lemma 4.5 and Lemma 4.6\" does not resolve this discrepancy. This is likely fixable, but it confirms that the Proposition A.2 application needs to be carried out in detail.","section":"§6.1, Eq. (6.1)"},{"comment":"The manuscript does not state a local well-posedness theorem for the perturbation system (1.2)–(1.3) in H². Theorem 1.1 asserts that \"the solution\" is global, and the bootstrap in Section 2.3 assumes a solution exists on [0,T]. Without an explicit LWP statement (or a reference), the global existence claim lacks a foundation. Please state the local well-posedness result for H² initial data, or give a precise reference, and explain how the bootstrap implies global existence via a maximal time argument.","section":"Theorem 1.1 and §2.3"},{"comment":"The bootstrap closure repeatedly says \"taking ε0 small enough\" in each proposition without specifying the ordering of the choices of the small constants. In particular, the imported estimates Propositions A.3–A.5 require a smallness condition on ∥du_{1,0}∥_{H⁴}+ν^{-1}∥∂t du_{1,0}∥_{H²}, which is controlled by E1. The constants C in the estimates may depend on the constants δ2,δ3,δ4 in those hypotheses, and on a,b. To make the closure rigorous, the paper should present an explicit ordering: first fix δ1,...,δ4 sufficiently small, then choose ε0 sufficiently small relative to those constants, then choose ε in Theorem 1.1. As written, the dependencies are not tracked, and it is not immediate that a single ε0 satisfies all the required smallness conditions simultaneously.","section":"§2.3 and Propositions 5.1–6.3"}],"minor_comments":[{"comment":"The symbol κ is used in Lemma 3.3 before its definition in (6.18). Please move the definition earlier or add a forward reference to avoid confusion.","section":"Section 6, Eq. (6.18) vs Lemma 3.3"},{"comment":"The factors in (4.16) and (4.33) use different constants, e^{(b-a)ν^{1/3}t} versus e^{(b-a)/2 ν^{1/3}t}. Both are plausible since b>a, but the inconsistency is distracting and should be harmonized.","section":"Lemma 4.7, Eq. (4.33) and Lemma 4.5, Eq. (4.16)"},{"comment":"The bullet says E4,1 is \"more simplified than that in [35], since there is an additional term ν^{2/3}∥△u3,≠∥_{X_b}\", but the definition of E4,1 does not include that term. This appears to be a leftover from an earlier formulation; please clarify the intended comparison.","section":"Remark 2.1, bullet for E4,1"},{"comment":"The abstract states the result for the Boussinesq system without specifying ν=μ or g=1, which are introduced in Section 1. Please state these assumptions in Theorem 1.1 or in the abstract for accuracy.","section":"Abstract and Theorem 1.1"},{"comment":"The paper repeatedly uses the phrase \"taking ε0 small enough\" and \"Cε0<1/2\" in several propositions. While this is standard in bootstrap proofs, a short summary of the bootstrap constants and their dependencies at the end of Section 2 would greatly improve readability.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is potentially publishable in a PDE journal if the omitted proof of Proposition A.2 is supplied and the technical discrepancies in the closure are fixed. The main theorem is significant, and the overall strategy is coherent. I do not see grounds to doubt the central claim, but the manuscript as submitted is not self-contained at a load-bearing point. The authors should be asked to provide a complete derivation of Proposition A.2, to resolve the X_a/X_b mismatch in (6.1), and to make the bootstrap ordering explicit. I would be willing to review a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper fills a real gap. For the 3D Boussinesq system with constant background temperature near Couette, earlier symmetrization tricks fail, and no Sobolev-space threshold was known. The authors prove that H² perturbations of size εν in velocity and εν² in temperature stay global, matching the 3D Navier-Stokes exponent for velocity and giving a plausible, new exponent for temperature. The quasi-linear decomposition Q = Q1 + νQ2 + Q3, with the good unknowns and the treatment of lift-up, is the right adaptation of Wei–Zhang’s machinery, and the bootstrap structure is coherent: the energy norms are balanced so that all the powers of ν actually close. The paper also gives a thoughtful heuristic optimality argument in Remark 1.2, clearly labeled as such.\n\nThe central soft spot is Proposition A.2, the coupled linear estimate for the pair (△u2,≠, u3,≠). It is stated without proof, with only the line “can be derived from Proposition 4.3 in [35], and we omit it.” That is not a cosmetic omission. This estimate carries the nonlocal lift-up coupling and the temperature forcing, and every later bound — E4,1, then E1, E2, E5, E6 — leans on it. The application in (6.1) also shows a small mismatch: the RHS uses ν^{-4/3} of (∂x², ∂z²)Θ≠ in Xb, while E5 only gives ∂z²Θ≠ in Xa. That is probably a typo or a fixable exponent adjustment, but it says the authors have not re-checked every detail of the imported estimate. Local well-posedness in H² is not stated either, which is minor for this community but should be included with the global theorem.\n\nNone of this is a demonstrated contradiction. The argument as a whole is internally consistent, the nonlinear estimates are long but carefully organized, and the citation pattern is clean. The paper deserves a serious referee, but the referee should be asked to verify Proposition A.2 in full, or the authors should supply an appendix proving it. I would recommend sending it to peer review with that request rather than desk-rejecting.","headline":"First global Sobolev stability threshold for the unstratified 3D Boussinesq near Couette, with the velocity at ν and temperature at ν², but the proof rests on an unproved imported linear estimate that needs a complete derivation before the result can be certified.","tokens_in":45941,"tokens_out":1923,"would_cite":true,"duration_ms":20473,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","76E05","35B35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that 3D Boussinesq perturbations of Couette flow with velocity size $\\varepsilon\\nu$ and temperature size $\\varepsilon\\nu^2$ in $H^2$ remain global in time.","keywords":["3D Boussinesq system","Couette flow","stability threshold","enhanced dissipation","lift-up effect","global regularity","Sobolev spaces","quasi-linearization"],"falsifier":"Solve the linearized Boussinesq equations around the modified Couette profile $V=y+d u_{1,0}$ with a nonzero-mode temperature forcing and check numerically whether the time-weighted $X_a$ and $X_b$ bounds of Propositions A.1–A.5 hold with the stated powers of $\\nu$; a single mode whose enhanced-dissipation exponent is worse than $\\nu^{1/3}$, or a zero-mode contribution that violates the $P_0$ conditions, would break the bootstrap.","tokens_in":44913,"feed_emoji":"🌊","tokens_out":9695,"duration_ms":78147,"temperature":0.7,"pith_summary":"This paper proves a stability threshold for the three-dimensional Boussinesq system near Couette flow in the unstratified, constant-temperature case, where the 3D lift-up mechanism is active rather than suppressed by stratification. The theorem states that if the initial velocity perturbation is at most $\\varepsilon\\nu$ and the initial temperature perturbation at most $\\varepsilon\\nu^2$ in $H^2$, with $\\varepsilon$ independent of the Reynolds number, then the solution exists globally in time. This gives Sobolev threshold exponents $\\beta=1$ for velocity and $\\beta=2$ for temperature, matching the formal balance of the lift-up and thermal-diffusion effects. A sympathetic reader should care because it extends the sharp Sobolev threshold known for 3D Navier-Stokes to the Boussinesq system while showing that temperature must be initialized one power of $\\nu$ smaller than velocity.","feed_headline":"Velocity ν, temperature ν²: Couette flow stays stable","feed_subtitle":"An H² proof shows the unstratified 3D Boussinesq system remains global whenever initial data meet those two size conditions.","key_machinery":"The argument is carried by a six-component energy functional ($E_1,\\dots,E_6$) with time weights $e^{a\\nu^{1/3}t}$, measuring zero and nonzero Fourier modes separately. Two devices make the closure possible: the split $u_{1,0}=d u_{1,0}+g u_{1,0}$ isolates the part of the streamwise zero mode that suffers lift-up amplification, and the good unknown $Q=u_{2,\\neq}+\\kappa u_{3,\\neq}$, with $\\kappa=\\partial_z V/\\partial_y V$ and $V=y+d u_{1,0}$, absorbs the worst velocity–temperature coupling terms. The proof imports linear space-time estimates for the moving-background operator $L_V=\\partial_t-\\nu\\Delta+V\\partial_x$ (Propositions A.1–A.5), which supply the enhanced dissipation and inviscid damping rates on which the nonlinear bootstrap rests.","core_discovery":"The central discovery is Theorem 1.1: for the perturbation system (1.2)–(1.3) on $\\mathbb{T}\\times\\mathbb{R}\\times\\mathbb{T}$ with equal viscosity and thermal diffusivity $\\nu$, any data with $\\|u_{\\rm in}\\|_{H^2}\\leq\\varepsilon\\nu$ and $\\|\\Theta_{\\rm in}\\|_{H^2}\\leq\\varepsilon\\nu^2$ produce a global solution, provided $\\varepsilon$ is sufficiently small and independent of $\\nu$. The proof handles the genuinely unstratified case $\\alpha=0$, where the standard symmetrization of stratified Boussinesq fails, and controls the 3D lift-up effect by introducing new good unknowns and a quasi-linear decomposition. The authors further argue from zero-mode balance that the two exponents are optimal: a temperature perturbation of order $\\nu^2$ is the largest that the linear transfer through $u_{2,0}$ can absorb without destabilizing.","pith_inferences":["The proof treats $\\nu=\\mu$ only; a natural extension would allow independent viscosity and thermal diffusivity and ask how the temperature exponent $\\beta=2$ changes with the ratio $\\mu/\\nu$.","The $H^2$ regularity is probably not the true boundary of the method; the same quasi-linear decomposition may push the threshold to lower Sobolev or Besov regularity, or to Gevrey data with a different exponent.","The imported linear estimates were proved for the Navier-Stokes operator; one can test their validity for the Boussinesq coupling by computing the spectrum or resolvent of the linearized operator around the modified profile $V$ including the $\\Theta$ feedback.","The same good-unknown construction may transfer to other shear flows with a lift-up instability, such as Poiseuille or Kolmogorov flow, whenever the background flow admits the $\\kappa$-weighted derivative structure."],"forward_implications":["Global-in-time existence follows for all $H^2$ data meeting the two size conditions, so no finite-time singularity develops near Couette flow at these amplitudes.","The threshold exponents match what formal asymptotics predict: $\\beta=1$ for velocity and $\\beta=2$ for temperature, with Remark 1.2 explaining why larger temperature data should destabilize through the zero-mode channel.","When the temperature is set to zero the result reduces to the known Sobolev threshold for 3D Navier-Stokes, so the theorem is consistent with the purely hydrodynamic case.","The weighted norms force nonzero modes to decay on the enhanced-dissipation time scale $\\nu^{-1/3}$, while the zero mode relaxes toward a modified Couette profile.","The result covers the unstratified case $\\alpha=0$, complementing the stratified case where the lift-up effect is suppressed and a smaller threshold exponent is available."],"supporting_citations":[{"why":"Supplies the imported linear space-time estimates (Propositions A.1, A.3–A.5 and Lemma A.1) and the quasi-linear good-unknown technique on which the energy closure is built.","marker":"[35]"},{"why":"Frames the transition-threshold problem and establishes the Sobolev threshold for 3D Navier-Stokes that the present result extends to Boussinesq.","marker":"[3]"},{"why":"Defines the stratified 3D Boussinesq benchmark with threshold $\\beta=11/12$; the paper targets its $\\alpha=0$ unstratified limit where symmetrization fails.","marker":"[14]"},{"why":"Provides the zero-mode decomposition $u_{1,0}=d u_{1,0}+g u_{1,0}$ that isolates the lift-up-affected part of the velocity.","marker":"[11]"},{"why":"A formal asymptotic analysis identifying $\\beta=1$ for streamwise and oblique perturbations, cited as evidence that the velocity exponent is optimal.","marker":"[8]"},{"why":"Establishes linear enhanced dissipation for 3D Boussinesq around stably stratified Couette, the background against which the unstratified case is contrasted.","marker":"[13]"}],"fun_headline_variants":["Optimal ν, ν² thresholds: 3D Boussinesq Couette flow stable","Global Boussinesq solutions at ν velocity, ν² temperature","Stability threshold: ν for velocity, ν² for heat in 3D Couette-Boussinesq","ν and ν² scaling: exact stability threshold in Boussinesq"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes, without reproving, that the imported linear decay estimates for the moving background $L_V$ remain valid for all the velocity–temperature coupling terms that appear in the Boussinesq equations, under the smallness condition on $d u_{1,0}$ that the bootstrap itself maintains.","fun_headline_variants_meta":{"raw":{"variants":["Optimal ν, ν² thresholds: 3D Boussinesq Couette flow stable","Global Boussinesq solutions at ν velocity, ν² temperature","Stability threshold: ν for velocity, ν² for heat in 3D Couette-Boussinesq","ν and ν² scaling: exact stability threshold in Boussinesq"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001298,"raw_usage":{"total_tokens":5264,"prompt_tokens":880,"completion_tokens":4384,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":496,"completion_tokens_details":{"reasoning_tokens":4293}},"tokens_in":496,"tokens_out":4384,"duration_ms":28420,"temperature":1.0,"reasoning_tokens":4293,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:04:11.567938+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the linearized Boussinesq equations around the modified Couette profile $V=y+d u_{1,0}$ with a nonzero-mode temperature forcing and check numerically whether the time-weighted $X_a$ and $X_b$ bounds of Propositions A.1–A.5 hold with the stated powers of $\\nu$; a single mode whose enhanced-dissipation exponent is worse than $\\nu^{1/3}$, or a zero-mode contribution that violates the $P_0$ conditions, would break the bootstrap.","supporting_citations":[{"cited_title":"and Zhang Z","cited_arxiv_id":null,"evidence_quote":"Supplies the imported linear space-time estimates (Propositions A.1, A.3–A.5 and Lemma A.1) and the quasi-linear good-unknown technique on which the energy closure is built."},{"cited_title":"and Masmoudi N","cited_arxiv_id":null,"evidence_quote":"Frames the transition-threshold problem and establishes the Sobolev threshold for 3D Navier-Stokes that the present result extends to Boussinesq."},{"cited_title":"and Zhang Z","cited_arxiv_id":null,"evidence_quote":"Provides the zero-mode decomposition $u_{1,0}=d u_{1,0}+g u_{1,0}$ that isolates the lift-up-affected part of the velocity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"A formal asymptotic analysis identifying $\\beta=1$ for streamwise and oblique perturbations, cited as evidence that the velocity exponent is optimal."},{"cited_title":"and Del Zotto A","cited_arxiv_id":null,"evidence_quote":"Establishes linear enhanced dissipation for 3D Boussinesq around stably stratified Couette, the background against which the unstratified case is contrasted."}],"review_version":1}