{"id":"95d2ecaf-b1e5-4f03-9afe-b4d23be020ac","arxiv_id":"2506.03679","paper_version":1,"verdict":"ACCEPT","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The stability threshold α=1/3 for 2D Boussinesq-Couette flow holds for unequal viscosity and thermal diffusivity, with H^{s+1/2} regularity for s>3/2.","lead":"This paper proves that small perturbations of Couette flow in a stratified 2D Boussinesq system stay stable when the initial disturbance is below a sharp size, roughly the cube root of the smallest diffusivity. It extends the known one-third stability threshold from equal to unequal viscosity and thermal diffusivity, and lowers the required regularity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.1 is misstated and its proof does not yield the convolution bound used in Lemma 2.2 and Lemma 5.3, so the nonlinear energy closure is not justified as written.","rationale":"The reader identified the ratio condition on ν,μ as the weakest assumption and accepted the paper. However, a more fundamental issue appears in Lemma 2.1: the proof does not establish the stated inequality, and the stated inequality does not match the form used in Lemma 5.3 and Lemma 2.2. Because Lemma 2.1 underpins the L1 estimate for u2 and the Υ-based estimates in the long-time nonlinear analysis, Proposition 6.1 lacks a valid foundation as written. The theorem may still be true, since a corrected convolution bound with the reciprocal ratio is plausible for the parameter ranges a,b≥1 used in the paper, but that corrected lemma is not proved. Therefore the appropriate verdict is CONDITIONAL: accept only after Lemma 2.1 is corrected and the affected estimates (Lemma 2.2, Lemma 5.3, Proposition 6.1) are re-verified. This is a concrete proof gap independent of the explicitly stated physical-parameter condition.","tokens_in":41060,"tokens_out":43148,"duration_ms":413704,"concrete_test":"Re-derive the exact convolution inequality needed in Lemma 5.3 for a=⟨l⟩, b=|k−l|≥1, λ=δ: verify numerically for l,k∈[-20,20], z∈[-10,10], t up to 10^3 whether ∫ |b|^δ/(|a,η|^{1+δ}|b,z−η|^{1+δ})dη ≤ C a^{-δ}(a+b)^δ/|a+b,z|^{1+δ} holds with a universal C. Also check whether Lemma 2.1 as printed (with (ab)^δ/(a+b)^δ) is compatible with this bound; if not, correct Lemma 2.1 and re-check Lemma 2.2's L1 estimate and Proposition 6.1's closure.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Lemma 2.1 (Section 2) is the convolution estimate used in the L1 bound of Lemma 2.2 and in the Υ-term bound of Lemma 5.3, which are needed for the long-time nonlinear estimates in Proposition 6.1. The proof first shows I := ∫ dη/(|a,η|^{1+λ}|b,z−η|^{1+λ}) ≤ C|a+b|^λ/(|ab|^λ). It then says 'dividing both sides by |a+b,z|^{1+λ} gives (2.1)', but division gives I/|a+b,z|^{1+λ} ≤ C|a+b|^λ/(|ab|^λ|a+b,z|^{1+λ}), not I ≤ C|ab|^λ/(|a+b|^λ|a+b,z|^{1+λ}). Moreover, the inequality actually invoked in Lemma 5.3 is ∫ |b|^δ/(|a,η|^{1+δ}|b,z−η|^{1+δ}) dη ≤ C a^{-δ}(a+b)^δ/|a+b,z|^{1+δ}, which has the reciprocal ratio (a+b)^δ/(ab)^δ, not (ab)^δ/(a+b)^δ. Thus the stated Lemma 2.1 is inconsistent with its applications; the proof of Theorem 1.1 rests on an unjustified convolution bound. The ratio condition on ν,μ is a separate, explicitly stated assumption, not the main gap.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a transition threshold of α≤1/3 for the 2D Boussinesq equations around Couette flow in T×R, for Richardson number γ^2>1/4 and possibly different viscosity ν and thermal diffusivity μ. The main theorem, Theorem 1.1, asserts that if the initial perturbation in H^{s+1/2}, s>3/2, has norm at most cκ^{1/3} with κ=min{ν,μ}, and if (ν+μ)/(2γ√νμ)<2−ε, then the solution is global and satisfies inviscid damping and enhanced dissipation estimates. The proof is divided into a short-time regime t≤κ^{-1/6} and a long-time regime t≥κ^{-1/6}, using weighted Fourier multipliers A_k and M, a coupled positive energy, and a large amount of multiplier-difference and convolution analysis to control the nonlinear terms.","tokens_in":41309,"tokens_out":16972,"duration_ms":175704,"significance":"If correct, the result improves the previous α=1/2 threshold for Boussinesq–Couette stability to α=1/3, matching the known 2D Navier–Stokes threshold, and it removes the ν=μ restriction from recent work by Knobel while allowing lower Sobolev regularity. The paper is largely self-contained: the short- and long-time energy propositions are stated precisely, the multiplier lemmas are proved in detail, and the nonlinear estimates are reduced to symmetrization and convolution inequalities. However, a central convolution lemma is misstated and its proof is not valid in the range of parameters actually used; this is a load-bearing gap that is repairable but must be fixed before the theorem is fully supported.","major_comments":[{"comment":"Lemma 2.1 is stated with the reciprocal ratio: it claims ∫ dη/(|a,η|^{1+λ}|b,z−η|^{1+λ}) ≤ C |ab|^λ/(|a+b|^λ |a+b,z|^{1+λ}). The proof actually yields, after division, the opposite ratio C |a+b|^λ/(|ab|^λ |a+b,z|^{1+λ}). Moreover, the intermediate inequality |a|^λ+|b|^λ ≤ |a+b|^λ used in the proof is false for 0<λ<1, which is exactly the range λ=δ<1/2 needed in Lemma 5.3. Lemma 2.2 uses λ=2 and would work with the corrected ratio after cancellation of |k|^2, whereas with the stated ratio the resulting bound would grow like k^4 and the summation over k would fail. Lemma 5.3 explicitly requires the corrected ratio: the application with a=⟨l⟩, b=|k−l|, λ=δ gives ⟨l⟩^{-δ}(⟨l⟩+|k−l|)^δ / |⟨l⟩+|k−l|, ξ−(k−l)t|^{1+δ}, not what the stated (2.1) provides. Since Lemma 5.3 controls the Υ-term used in the long-time estimates of Proposition 6.1, the proof of Theorem 1.1 is not fully justified as written. The intended bound is elementary and the gap is repairable, but the statement, the proof, and the applications must be corrected consistently.","section":"Section 2 (Lemma 2.1), used in Lemmas 2.2 and 5.3"}],"minor_comments":[{"comment":"The statement that the regularity assumption 'should be sharp' is a heuristic comparison with a linear lower bound; no instability or optimality proof is supplied. Please rephrase as a consistency remark or provide a precise statement.","section":"Remark 1.1"},{"comment":"There is a typo 'A_k(t,ξ)∼∼' with a doubled tilde; it should read 'A_k(t,ξ)∼'.","section":"Section 3, beginning"},{"comment":"The condition (ν+μ)/(2γ√νμ)<2−ε is an additional restriction on the physical parameters, not a consequence of the Sobolev threshold. It is used to absorb the linear cross term I1; the paper should state this role explicitly and discuss, even briefly, how restrictive it is for γ near 1/2.","section":"Theorem 1.1 and Section 4, Eq. (4.13)"},{"comment":"In the Gronwall step, the denominator is written as 1−C_+(t+t^2)c1κ^{1/3}; since the differential inequality is dE/dt≤C⟨t⟩E^{3/2}, the exact integration produces factors of 1/2 or 1/4. These can be absorbed into C_+, but the displayed formula should be checked for consistency.","section":"Section 2, proof of Proposition 2.3"}],"recommendation":"major_revision","confidential_remarks":"The main theorem is plausible and the overall strategy is coherent, but the convolution lemma is load-bearing and currently both misstated and incorrectly proved for the parameter range used. This is not grounds for rejection because the corrected inequality is elementary and the rest of the proof appears consistent with it; however, the authors must fix the lemma and re-verify the constants in Lemma 5.3 and Proposition 6.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: this paper delivers a real improvement over the known threshold results. It proves the 1/3 stability threshold for 2D Boussinesq–Couette with unequal viscosity and thermal diffusivity, at H^{s+1/2} (s>3/2) regularity, where Knobel [29] had nu=mu in H^N and the authors' own earlier paper had only 1/2. The proof follows the Wei–Zhang two-timescale multiplier strategy and lays out the energy estimates in detail. I did not find a hidden circularity; the bootstrap uses the bootstrap assumption only via the quadratic energy upper bound, which is standard.\n\nThe stress-test note about Lemma 2.1 does not hold up. The printed inequality has the ratio reversed—it reads |ab|^lambda/(|a+b|^lambda) where the proof gives (|a+b|^lambda)/(|ab|^lambda). But the proof is correct for the reciprocal ratio, and that reciprocal ratio is exactly what is used in Lemma 2.2 and Lemma 5.3. So the lemma statement is a typo, not a gap. The convolution bounds used in the nonlinear estimate are justified by the proven version. This should be fixed in revision, but it does not undermine Theorem 1.1.\n\nThe main soft spot is that the proof is heavy; I gave the multiplier bounds a careful but not line-by-line check, and the constants look consistent. The ratio condition (nu+mu)/(2gamma sqrt(nu mu)) < 2-epsilon is a real restriction, but it is stated in the theorem and used explicitly to close the linear estimate; for equal diffusivities it is automatic, and for unequal ones it is a reasonable parameter constraint. The regularity claim and the remark on sharpness are plausible and well-grounded in the linear theory.\n\nThis paper is for the hydrodynamic stability community. It is a legitimate advance and deserves a serious referee. I would send it out and let the referee require a corrected Lemma 2.1 and some tidying of the multiplier estimates. My own verdict: accept after minor revision.","headline":"Solid extension of the 1/3 threshold to unequal diffusivities at low regularity; the flagged Lemma 2.1 gap is a typo, not a flaw.","tokens_in":41932,"tokens_out":5145,"would_cite":true,"duration_ms":48405,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","76E05","76D05"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that 2D Boussinesq Couette flow with Richardson number greater than 1/4 is asymptotically stable against Sobolev perturbations of size at most cκ^(1/3), establishing the transition threshold α = 1/3.","keywords":["Couette flow","Boussinesq equations","transition threshold","inviscid damping","enhanced dissipation","Richardson number","Sobolev stability","stratified shear flow"],"falsifier":"Compute the linearized Boussinesq evolution with ν = μ = 1 and γ = 1/2, the excluded case where the absorption coefficient in (4.13) equals 2; if the solution still exhibits the exp(−ϵκ^(1/3)t) decay of (1.7), the paper's central claim would be undercut, since the proof's key bound fails exactly there. Alternatively, a numerical simulation of the full nonlinear system at perturbation amplitude κ^(1/3) with (ν+μ)/(2γ√(νμ)) ≥ 2−ε and γ² > 1/4 that transitions to turbulence would falsify the threshold claim.","tokens_in":40816,"feed_emoji":"🌊","tokens_out":7138,"duration_ms":84325,"temperature":0.7,"pith_summary":"This paper establishes the transition threshold α = 1/3 for the 2D Boussinesq equations near Couette flow: if the initial perturbation has Sobolev size at most c(min{ν,μ})^(1/3), the flow stays global and returns to the laminar profile. This matches the optimal threshold known for the unstratified 2D Navier-Stokes equations, so the buoyancy coupling does not worsen the stability window when the Richardson number γ² > 1/4. The proof works for different viscosity ν and thermal diffusivity μ, provided their ratio satisfies (ν+μ)/(2γ√(νμ)) < 2−ε, and it requires only H^(s+1/2) initial data with s > 3/2. A sympathetic reader should care because this quantifies the maximal disturbance a stratified shear flow can absorb before transition, the modern PDE version of Reynolds' classical question.","feed_headline":"Stratified Couette flow resists perturbations up to (viscosity)^{1/3}","feed_subtitle":"The 1/3 threshold for 2D Navier-Stokes extends to Boussinesq flows when the Richardson number exceeds 1/4.","key_machinery":"The argument is carried by a frequency-adapted energy D(t) = ‖mû‖²_{L²} + γ²‖mϑ̂‖²_{L²} + Re∫(mϑ̂)(mû₁)dξ, which is positive exactly when γ > 1/2. The multiplier m(t,k,ξ) has two regimes: for t ≤ κ^(−1/6) it is |k₊, ξ−kt|^(1/2)⟨k,ξ⟩^s $e^{{M₀}}$, capturing inviscid damping; for t ≥ κ^(−1/6) it adds the factor $e^{{ϵκ^(1/3)t}}$ together with compound terms M₁+M₂+M₃ that suppress echo cascades and capture enhanced dissipation. The cross term Re(mϑ̂ mû₁) combined with the parameter condition (ν+μ)/(2γ√(νμ)) < 2−ε absorbs the linear buoyancy-velocity coupling into the diffusion terms, and symmetrized multiplier bounds reduce all nonlinear terms to cubes of the weighted energy, so that a Grönwall estimate closes on the short time scale and a bootstrap closes on the long time scale.","core_discovery":"Theorem 1.1 states that for κ = min{ν,μ} ∈ (0,1), γ > 1/2, s > 3/2, and (ν+μ)/(2γ√(νμ)) < 2−ε, there are constants c, ϵ depending only on γ, s, ε such that whenever ‖(u_in, ϑ_in)‖_{$H^{{s+1/2}}$} ≤ cκ^(1/3), the perturbation remains globally bounded and satisfies the inviscid damping estimate ‖(u1)≠‖_{L²} + ⟨t⟩‖u2‖_{L²} + ‖ϑ≠‖_{L²} ≤ C⟨t⟩^(−1/2)e^(−ϵκ^(1/3)t)‖(u_in, ϑ_in)‖ and the enhanced dissipation estimate ‖⟨∇_L⟩^(1/2)(u≠, ϑ≠)‖_{H^s} ≤ Ce^(−ϵκ^(1/3)t)‖(u_in,ϑ_in)‖. In other words, perturbations of size up to κ^(1/3) are absorbed by the Couette mixing mechanism, with the non-zero Fourier modes decaying on the enhanced time scale κ^(−1/3) rather than the ordinary viscous time scale κ^(−1). The regularity assumption H^(s+1/2) is argued to be sharp in the sense that even the linearized Euler-Boussinesq problem needs H² data for the corresponding inviscid damping decay.","pith_inferences":["The excluded equality case (ν+μ)/(2γ√(νμ)) = 2 is a natural boundary: the ε-dissipation in the energy estimate shrinks to zero there, so the threshold α = 1/3 may fail or need a different mechanism when viscosity and thermal diffusivity are extremely unequal.","At γ = 1/2, the cross-term energy becomes degenerate, so the true transition threshold may be different exactly at the Miles-Howard critical Richardson number; a separate analysis of the critical case would test whether α = 1/3 persists there.","The same two-time-scale multiplier and bootstrap strategy should transfer to other dissipative fluid systems with stable stratification or magnetic effects, where the analogous threshold for Couette flow is either open or known only under more restrictive parameter assumptions."],"forward_implications":["If the theorem is correct, stratified Couette flow with γ² > 1/4 has the same Sobolev transition threshold α = 1/3 as 2D Navier-Stokes, so buoyancy does not reduce the basin of attraction at fixed viscosity.","The non-zero spatial modes of velocity and temperature decay like e^(−ϵκ^(1/3)t), meaning the effective mixing time is O(κ^(−1/3) log(1/κ)), much faster than the diffusive time O(κ^(−1)).","The inviscid damping estimate ⟨t⟩‖u₂‖_{L²} ≤ C⟨t⟩^(−1/2)e^(−ϵκ^(1/3)t) implies the vertical velocity collapses algebraically in the Eulerian frame even while the enhanced dissipation kills the fluctuating part.","The result allows ν ≠ μ as long as (ν+μ)/(2γ√(νμ)) < 2−ε, removing the equal-diffusivity restriction of the recent H^N threshold result.","The proof only needs H^(s+1/2) initial data with s > 3/2, and the paper argues this regularity is sharp for the inviscid damping effect."],"supporting_citations":[{"why":"Supplies the two-time-scale multiplier and bootstrap method, plus the 1/3 threshold for 2D Navier-Stokes Couette flow that this paper adapts.","marker":"[48]"},{"why":"Provides the energy estimate used here to remove the ν = μ restriction and to work with H^{s+1/2} data.","marker":"[40]"},{"why":"Gives the prior 1/2 threshold for Boussinesq Couette flow with ν ∼ μ and large Richardson number, the baseline this paper improves.","marker":"[53]"},{"why":"Establishes linear inviscid damping for stratified Couette flow and the H² regularity requirement used to argue sharpness of the regularity assumption.","marker":"[51]"},{"why":"Very recently obtained the 1/3 threshold for ν = μ in H^N using symmetric variables, the concurrent result this paper extends to different diffusion coefficients and lower regularity.","marker":"[29]"},{"why":"One of the proofs of the 1/3 threshold for 2D Navier-Stokes Couette flow whose short/long time decomposition is referenced.","marker":"[37]"}],"fun_headline_variants":["Boussinesq Couette flow stable up to viscosity^1/3","1/3 threshold extends to stratified Couette flows","Sharp inviscid damping for Boussinesq Couette","Couette turbulence threshold holds under stratification","Richardson > 1/4: Couette flow resists perturbations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof needs the ratio (ν+μ)/(2γ√(νμ)) to stay strictly below 2; if that ratio equals or exceeds 2, the linear buoyancy-velocity coupling can no longer be absorbed by diffusion, and the energy estimate that closes the argument breaks down.","fun_headline_variants_meta":{"raw":{"variants":["Boussinesq Couette flow stable up to viscosity^1/3","1/3 threshold extends to stratified Couette flows","Sharp inviscid damping for Boussinesq Couette","Couette turbulence threshold holds under stratification","Richardson > 1/4: Couette flow resists perturbations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1376,"prompt_tokens":1064,"completion_tokens":312,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":680,"completion_tokens_details":{"reasoning_tokens":228}},"tokens_in":680,"tokens_out":312,"duration_ms":3934,"temperature":1.0,"reasoning_tokens":228,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:57:29.436502+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the linearized Boussinesq evolution with ν = μ = 1 and γ = 1/2, the excluded case where the absorption coefficient in (4.13) equals 2; if the solution still exhibits the exp(−ϵκ^(1/3)t) decay of (1.7), the paper's central claim would be undercut, since the proof's key bound fails exactly there. Alternatively, a numerical simulation of the full nonlinear system at perturbation amplitude κ^(1/3) with (ν+μ)/(2γ√(νμ)) ≥ 2−ε and γ² > 1/4 that transitions to turbulence would falsify the threshold claim.","supporting_citations":[{"cited_title":"Wei and Z","cited_arxiv_id":null,"evidence_quote":"Supplies the two-time-scale multiplier and bootstrap method, plus the 1/3 threshold for 2D Navier-Stokes Couette flow that this paper adapts."},{"cited_title":"Ren and D","cited_arxiv_id":null,"evidence_quote":"Provides the energy estimate used here to remove the ν = μ restriction and to work with H^{s+1/2} data."},{"cited_title":"Zhai and W","cited_arxiv_id":null,"evidence_quote":"Gives the prior 1/2 threshold for Boussinesq Couette flow with ν ∼ μ and large Richardson number, the baseline this paper improves."},{"cited_title":"Yang and Z","cited_arxiv_id":null,"evidence_quote":"Establishes linear inviscid damping for stratified Couette flow and the H² regularity requirement used to argue sharpness of the regularity assumption."},{"cited_title":"Suppression of Fluid Echoes and Sobolev Stability Threshold for 2D Dissipative Fluid Equations Around Couette Flow","cited_arxiv_id":"2505.23391","evidence_quote":"Very recently obtained the 1/3 threshold for ν = μ in H^N using symmetric variables, the concurrent result this paper extends to different diffusion coefficients and lower regularity."},{"cited_title":"Masmoudi and W","cited_arxiv_id":null,"evidence_quote":"One of the proofs of the 1/3 threshold for 2D Navier-Stokes Couette flow whose short/long time decomposition is referenced."}],"review_version":1}