{"id":"aaafaadc-1d92-4c66-8883-41d1f9a2512e","arxiv_id":"2505.08486","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Localized vorticity perturbations of 2D Couette flow relax to the linearized diffusion profile with rate t^{-1/2} after an explicit Reynolds-dependent time, for large data.","lead":"This mathematics paper proves that vorticity disturbances around the standard shear flow between two sliding plates eventually settle into a smooth, predictable diffusive shape, even when the disturbances start out large. It is the first quantitative large-data relaxation result for such flows and pins down the waiting time in terms of the Reynolds number.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Abstract overclaims: Theorem 1.3 requires Ω(1)∈L2(m0) with m0>7m/2; Theorem 1.1's L1 well-posedness does not imply this, and explicit L1 data can have a t=1 tail outside every such space.","rationale":"I read the paper as establishing a quantitative, large-data asymptotic stability theorem for the 2D Navier–Stokes vorticity equation near Couette flow, valid for rescaled data in the weighted space L2(m0) at time t=1. The main proof is elaborate but internally coherent: the rescaled equation (1.10) has a small nonlinear coefficient at large times, the energy estimates in Section 3 provide the needed L2(m) and derivative bounds, and the semigroup estimates in Section 4 yield the t^{-1/2} relaxation to the Gaussian profile. I did not find a clear internal contradiction in the proof of Theorem 1.3; the parameter bookkeeping for T1 and the final powers is consistent with the condition m0>7m/2. The weakest point is exactly the gap between the advertised scope and the theorem's hypothesis. The L1 global well-posedness of Theorem 1.1 does not imply polynomial weighted decay at time 1, and explicit L1 data with power-law tails remain power-law at t=1 under the linear evolution, putting them outside every L2(m0) with m0≥2. Thus the abstract's promise of 'general perturbations, including point vortices' is not supported by any stated theorem; the paper itself acknowledges measure-valued data and general L1 asymptotics as open problems. This is the same concern the reader identified, so I agree with the conditional verdict. Since no new internal flaw was found, I recommend the reader's verdict remain unchanged: the stated theorems can be accepted conditionally, provided the abstract and marketing statements are revised to match the actual L2(m0) hypothesis.","tokens_in":51010,"tokens_out":33577,"duration_ms":311508,"concrete_test":"Compute analytically or numerically the time-one linear evolution S(1)ω0 for ω0(x,y)=(1+|x|^2+|y|^2)^{-3/2} using the explicit kernel (1.4) with ν=1, and express the result in the rescaled variables (1.8)–(1.9) at t=1. If the profile has the expected tail |z|^{-3} (equivalently ⟨X,Y⟩^{-3}), then the weighted integral ∫|Ω(1)|^2⟨X,Y⟩^{2m0}dXdY diverges for every m0≥2, so the hypothesis m0>7m/2 of Theorem 1.3 fails for this L1 datum. This settles whether the L2(m0) assumption can be inferred from the L1 well-posedness theory: it cannot. For the nonlinear claim, repeat the check with a small multiple εω0: the nonlinear correction is O(ε^2) and cannot remove the leading ε|z|^{-3} tail, so the same failure persists.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central advertised statement is the abstract's asymptotic-convergence claim for 'general perturbations... including singular configurations such as point vortices.' The only quantitative theorem supporting this is Theorem 1.3, whose hypothesis is Ω(1)∈L2(m0) with m0>7m/2 (m≥3) in the rescaled coordinates (1.8)–(1.9). This hypothesis is not a consequence of Theorem 1.1: the L1 theory only yields ω(1)∈L∞ and Lp bounds, with no spatial decay. The gap is real: take f(z)=(1+|z|^2)^{-3/2}∈L1(R2). For the linear semigroup S(1) with kernel (1.4), the kernel is a nondegenerate Gaussian at t=1, so (S(1)f)(z)∼|z|^{-3} as |z|→∞. In the rescaled variables this means Ω(1;X,Y)∼⟨X,Y⟩^{-3}, which lies in L2(m0) only for m0<2, hence is outside every space L2(m0) with m0≥3, a fortiori m0>7m/2. Thus Theorem 1.3 does not cover general L1 perturbations, and the point-vortex case is explicitly left open in Sections 1.2 and Remark 1.6. This is an overclaim in the abstract rather than an internal inconsistency of the stated theorem, but it is load-bearing for the advertised scope.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the 2D Navier-Stokes equations near planar Couette flow in the whole plane. It first establishes global well-posedness of the vorticity formulation for arbitrary L^1 initial data (Theorem 1.1), with heat-like L^p decay. For the rescaled vorticity in variables adapted to the explicit fundamental solution G_L, it proves weighted L^2(m) growth estimates and higher-derivative bounds (Theorem 1.2), and then shows that, for initial data at time t=1 in a sufficiently strong weighted space, the vorticity converges to the mass times the Gaussian fundamental solution of the linearized problem at rate t^{-1/2}, after an explicit but very large time T_1 (Theorem 1.3). The proof combines a change of variables based on G_L, hypocoercive energy estimates with logarithmic coefficients, and an explicit Fourier solution formula for the limiting Fokker-Planck operator L_∞. The abstract advertises coverage of general perturbations 'including singular configurations such as point vortices'; this advertised scope is not supported by the stated theorems.","tokens_in":51252,"tokens_out":13052,"duration_ms":133623,"significance":"If the main theorem is correct as stated, it constitutes a quantitative, large-data asymptotic stability result for Couette flow in R^2, with no smallness assumption on the perturbation and with an explicit (though extremely large) waiting time. Notable strengths include the explicit Fourier formula for the semigroup generated by L_∞, the clean derivation of the decay rate e^{-\\tau/2} in weighted spaces, and the transparent parameter dependence in the bounds. The paper also gives a useful comparison with the transition-threshold literature and with the prior work of Gallay-Wayne. However, the actual theorems require the rescaled vorticity at t=1 to belong to L^2(m_0) with m_0 > 7m/2; this is a genuine spectral/decay condition that is not implied by the L^1 well-posedness theory. The advertised inclusion of point vortices is explicitly left open in Sections 1.2 and Remark 1.6, so the abstract overstates the scope of the results.","major_comments":[{"comment":"The abstract claims that the asymptotic convergence holds for 'general perturbations, which may be large and of low regularity, including singular configurations such as point vortices.' This claim is not a consequence of any theorem in the paper. Theorem 1.3 requires \\Omega(1)\\in L^2(m_0) with m_0 > 7m/2 (m\\ge 3), and the point-vortex case is explicitly identified as an open problem in §1.2 and Remark 1.6. Moreover, the L^1 well-posedness of Theorem 1.1 does not imply the weighted hypothesis: for example, f(z)=(1+|z|^2)^{-3/2}\\in L^1(R^2) evolves to a profile with a |z|^{-3} tail at t=1 under the linear semigroup, so the rescaled function lies in L^2(m_0) only for m_0<2, not for m_0\\ge 3. The abstract and introduction should be revised to state the actual hypothesis \\Omega(1)\\in L^2(m_0), and the point-vortex claim should be removed or clearly qualified as an open problem.","section":"Abstract and §1.2, Remark 1.6"},{"comment":"The proofs of Theorems 1.2 and 1.3 are based on a priori estimates for smooth solutions of (1.10), as stated at the start of Section 3, but no approximation or regularization argument is provided to justify that the mild solution constructed in Theorem 1.1 (with only L^1 initial data) satisfies these estimates. Since the paper presents these as unconditional theorems for the unique solution, the authors should either prove that the solution is smooth for positive times (which is plausible by parabolic smoothing and the Biot-Savart law) or add a standard density/approximation argument with uniform constants. As written, this is a gap in the proof of the central convergence theorem.","section":"Section 3, beginning and Proposition 3.1"}],"minor_comments":[{"comment":"The text says the L^1 norm is 'non-decreasing' with respect to time, but Lemma 2.2 and the surrounding discussion prove that it is non-increasing; the word should be corrected.","section":"Section 2.2, before Lemma 2.2"},{"comment":"There is a typo: 'Coutte flow' should be 'Couette flow'.","section":"Section 1.2, first paragraph"},{"comment":"The constant C in (1.7) appears to depend on p, since the proof via dyadic iteration produces p-dependent constants; the statement should explicitly say C=C(p) to avoid confusion.","section":"Theorem 1.1, estimate (1.7)"},{"comment":"The phrase 'the other powers are all negative' is somewhat misleading: the exponents in (4.20) and (4.21) contain positive powers of t_0 in the displayed constants, and their decay in t_0 comes from the chosen small parameters; this is presumably correct but the phrasing should be clarified.","section":"Section 4.2, after (4.25)"},{"comment":"The numerical value 'nearly T_2 \\approx (1+\\nu^{-1}\\|\\Omega(1)\\|)^{62}' is a useful illustration, but the phrase 'nearly' is informal; it would be clearer to state that the exponent is obtained in the limit m=3, m_0\\to\\infty, and that all terms depending on \\delta,\\sigma,\\epsilon are absorbed into the stated 'nearly'.","section":"Remark 1.5"}],"recommendation":"major_revision","confidential_remarks":"The paper contains a solid theorem-level architecture and the main quantitative convergence result appears defensible. However, the abstract's advertised scope—general low-regularity perturbations including point vortices—is contradicted by the actual hypotheses of Theorem 1.3 and by the paper's own open-problems section. I recommend that the editor ask the authors to realign the abstract and introduction with the proved theorems, and to add the missing smoothing/approximation justification in Section 3. The paper would then be a valuable contribution to the quantitative stability literature for Couette flow."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the main theorem is a real advance and is not broken, but the abstract overstates what is proved. Theorem 1.3 gives a quantitative, large-data convergence rate t^{-1/2} to the linear Couette kernel, with an explicit Reynolds-dependent waiting time. That is new and significant—it extends Gallay-Wayne to the Couette setting and makes the asymptotic quantitative. The construction in (1.8)-(1.9), the observation that the nonlinearity gains a factor nu^{-1}<t>^{-2}, and the explicit Fourier solution for the L_infinity semigroup in Proposition 4.1 are the core contributions. The energy estimates in Section 3 are heavy but laid out carefully. Credit where due: the paper does not just claim; it proves.\n\nThe soft spot is scoping. The abstract advertises 'general perturbations ... including singular configurations such as point vortices,' but the theorem needs the rescaled data at t=1 to lie in L^2(m0) with m0 > 7m/2. The L^1 well-posedness in Theorem 1.1 does not imply that. The stress-test example is correct: a function like (1+|z|^2)^{-3/2} in L^1 evolves under the Couette semigroup to a t=1 tail that lies in L^2(m0) only for m0 < 2, so outside every admissible space. Point vortices are explicitly left open in Section 1.2 and Remark 1.6. So the abstract's advertised scope is load-bearing and not supported by the stated theorems. This is fixable: revise the abstract or add a bridge result for measure-valued or weaker data. It is not an internal contradiction in Theorem 1.3, which appears coherent on its own terms.\n\nTwo minor issues: the proof of Lemma 2.2 has a small ambiguity about which velocity field drives the decomposed positive/negative equations; and the waiting time T2 in Remark 1.5 is enormous (power roughly 62), which the authors honestly flag as not sharp. Neither affects the main theorem.\n\nThis paper is for people working on asymptotic stability near shear flows and vortex relaxation. A serious referee should engage with it. My recommendation: send it to peer review. Expect heavy but doable revisions—mainly tightening the abstract to match the theorems. If the authors fix that, this could be a clean accept.","headline":"Real quantitative convergence theorem near Couette, but the abstract promises point-vortex and L1 coverage that the theorem hypotheses do not deliver.","tokens_in":51861,"tokens_out":2865,"would_cite":true,"duration_ms":29621,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q30","35B40","35B35","76D05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Large vorticity perturbations near planar Couette flow return to the linear diffusive profile at the optimal $t^{-1/2}$ rate, after a long but explicit time depending on the relative Reynolds number.","keywords":["Couette flow","Navier-Stokes equations","vorticity","long-time asymptotics","large data","Fokker-Planck semigroup","enhanced dissipation","inviscid damping"],"falsifier":"Fix any positive viscosity (for instance $\\nu=1$) and choose a datum $\\Omega(1)\\in L^2(m_0)$ satisfying the theorem with a very large relative Reynolds number $1+\\nu^{-1}\\|\\Omega(1)\\|_{L^2(m_0)}$. If an accurate numerical integration of the perturbed vorticity equation shows that $\\sup_{t\\ge T_1} t^{1/2}\\|\\Omega(t)-M(\\Omega(1))G\\|_{L^2(m)}$ grows without bound as the datum size increases, or that a two-vortex initial profile leaves a permanent two-hump residue rather than collapsing to a single Gaussian, then (1.17) and Remark 1.4 would be false.","tokens_in":50730,"feed_emoji":"🌀","tokens_out":10759,"duration_ms":100640,"temperature":0.7,"pith_summary":"The paper claims that the vorticity perturbation of the two-dimensional Navier-Stokes equations near planar Couette flow, starting from large and only mildly regular data, converges to a universal long-time profile: a constant multiple of the fundamental solution $G_L(t)$ of the linearized vorticity equation, with the optimal decay rate $t^{-1/2}$ in $L^1$. The convergence is quantitative and occurs after an explicit time $T_1$ that depends only on the relative Reynolds number $1+\\nu^{-1}\\|\\Omega(1)\\|_{L^2(m_0)}$; this is what allows the initial perturbation to be large. The proof is built on a new set of long-time coordinates adapted to the exact Couette kernel, in which the nonlinearity acquires a $\\langle t\\rangle^{-2}$ smallness even though the data are not small. The advertised singular cases such as point vortices are not covered by the stated theorems, which require the rescaled data at $t=1$ to lie in $L^2(m_0)$ with $m_0>7m/2$.","feed_headline":"Large vorticity near Couette flow relaxes to a Gaussian at t^-1/2","feed_subtitle":"New proof covers large, low-regularity data with an explicit waiting time set by the Reynolds number.","key_machinery":"The load-bearing construction is a time-dependent change of variables (1.8)-(1.9) built from the exact fundamental solution $G_L(t)$ of the linearized equation (1.3). It rescales the sheared coordinates by the spreading rates of $G_L$: $X=x/\\sqrt{\\nu t(1+t^2/3)}$ and $Y=((1+t^2/3)y-(t/2)x)/\\sqrt{\\nu t(1+t^2/3)(1+t^2/12)}$, and writes $\\omega(t,x,y)=\\Omega(t,X,Y)/(\\nu t\\sqrt{1+t^2/12})$. In these coordinates the equation becomes $t\\partial_t\\Omega=L_t\\Omega+N_t\\Omega$, where $L_t$ converges to the Fokker-Planck operator $L_\\infty=4\\partial_Y^2+2Y\\partial_Y+2+\\frac{\\sqrt{3}}{2}(X\\partial_Y-Y\\partial_X)$ whose Gaussian steady state is $G$, and the nonlinearity $N_t$ carries a prefactor $\\nu^{-1}(1+t^2/12)^{-1}$, hence decays like $t^{-2}$. The proof combines an explicit Fourier characteristic formula for $e^{\\tau L_\\infty}$ (which gives exponential decay on weighted $L^2(m)$) with a hypoelliptic energy functional whose successively ordered logarithmic weights propagate $\\partial_X$ and $\\partial_Y$ derivatives.","core_discovery":"On the paper's own terms, the main discovery is Theorem 1.3: for $3\\le m$, $m_0>\\frac{7}{2}m$, and $\\Omega(1)\\in L^2(m_0)$, the rescaled vorticity satisfies $\\|\\Omega(t)-M(\\Omega(1))G\\|_{L^2(m)}\\le C t^{-1/2}(1+\\nu^{-1}\\|\\Omega(1)\\|_{L^2(m_0)})^{\\frac{8(m+1)}{1-7m/2m_0}-1+\\epsilon}\\|\\Omega(1)\\|_{L^2(m_0)}$ for all $t\\ge T_1$, where $T_1$ has the same structure with exponent $\\frac{7(m+1)}{1-7m/2m_0}+\\epsilon$. In the original variables this gives $\\|\\omega(t)-M(\\omega(1))G_L(t)\\|_{L^1}=O(t^{-1/2})$ (Remark 1.4), so the nonlinear solution returns to the self-similar diffusive profile of the linearized problem, carrying the same total vorticity mass. The argument is not perturbative: the theorem holds for large data, and the price paid is the long explicit waiting time $T_1$ and the polynomial weight requirement $L^2(m_0)$.","pith_inferences":["The self-similar variables are built from the exact linear kernel $G_L$, so the same strategy may extend to measure-valued initial vorticity (such as a point vortex) once one shows such a solution enters $L^2(m_0)$ after a finite positive time; the paper itself leaves this as an open problem.","The exponents in $T_1$ are far from sharp; for $m=3$ and large $m_0$ they give a timescale of order $(1+\\nu^{-1}\\|\\Omega(1)\\|)^{62}$, which suggests numerical experiments on high-Reynolds Couette flow could test whether convergence is actually much faster.","One could try to replace the polynomial weight $L^2(m_0)$ by a tail smallness condition on $\\Omega(1)$; if that succeeds, the theorem would cover genuinely $L^1$ data that only develop weighted integrability after a short time.","The explicit Fourier representation of $e^{\\tau L_\\infty}$ is a transferable tool: analogous characteristics could give quantitative convergence to self-similar profiles for other shear flows with explicit linear kernels."],"forward_implications":["The nonlinear vorticity near Couette flow therefore relaxes, in $L^1$, to $M(\\omega(1))G_L(t)$ at the same $t^{-1/2}$ rate as the linearized flow, so the background shear does not change the universal late-time Gaussian character.","The waiting time before the $t^{-1/2}$ regime is explicit and finite, so the result is a large-data asymptotic statement rather than a limiting one: convergence is guaranteed after a computable time $T_1$ depending on the relative Reynolds number.","Because the limit profile $G_L$ itself exhibits enhanced dissipation and inviscid damping, the theorem implies these mechanisms prevail in the final stage even for large data.","Imposing the extra smallness $\\nu^{-1}\\|\\Omega(1)\\|_{L^2(m_0)}\\lesssim 1$ reproduces the transition-threshold condition $\\|\\omega(1)\\|_{L^2}\\lesssim \\nu^{1/2}$, matching the known $\\beta=1/2$ threshold in nearly $L^2$ norms.","Theorem 1.1 supplies the companion existence theory: every $L^1$ vorticity datum has a unique global mild solution, with $L^p$ decay $\\|\\omega(t)\\|_{L^p}\\lesssim(\\nu t)^{1/p-1}\\|\\omega_0\\|_{L^1}$."],"supporting_citations":[{"why":"establishes global solvability of the two-dimensional Navier-Stokes vorticity equation with $L^1$ data, used as the basis for Theorem 1.1.","marker":"[8]"},{"why":"supplies the uniqueness argument that fixes the $L^1$ mild solution of the integrated vorticity equation.","marker":"[10]"},{"why":"provides the model vortex-asymptotics result for 2-D Navier-Stokes that this paper extends to a Couette background.","marker":"[21]"},{"why":"supplies the spectral theory of Fokker-Planck operators used for the weighted semigroup estimates.","marker":"[24]"},{"why":"is the hypoellipticity tool used to gain horizontal derivatives in the energy estimate.","marker":"[26]"},{"why":"provides the measure-valued well-posedness context and the Nash-Moser iteration estimate used in Section 2.","marker":"[30]"},{"why":"fixes the sharp small-data transition threshold for Couette flow on the whole plane, which the authors match via their $\\nu$-smallness condition.","marker":"[33]"}],"fun_headline_variants":["Couette flow: large vorticity relaxes to Gaussian at t^-1/2","Nonperturbative: Couette flow forces Gaussian relaxation","Large, rough vorticity still yields Gaussian profile under Couette flow","Point vortices included: Couette flow forces Gaussian diffusive limit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the rescaled vorticity at time $t=1$ is square-integrable with enough polynomial spatial decay, specifically $\\Omega(1)\\in L^2(m_0)$ with $m_0>7m/2$ for some $m\\ge 3$; without that, the quantitative $O(t^{-1/2})$ convergence in the paper is not proved.","fun_headline_variants_meta":{"raw":{"variants":["Couette flow: large vorticity relaxes to Gaussian at t^-1/2","Nonperturbative: Couette flow forces Gaussian relaxation","Large, rough vorticity still yields Gaussian profile under Couette flow","Point vortices included: Couette flow forces Gaussian diffusive limit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000932,"raw_usage":{"total_tokens":3970,"prompt_tokens":906,"completion_tokens":3064,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":2987}},"tokens_in":522,"tokens_out":3064,"duration_ms":21718,"temperature":1.0,"reasoning_tokens":2987,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:54:54.828586+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix any positive viscosity (for instance $\\nu=1$) and choose a datum $\\Omega(1)\\in L^2(m_0)$ satisfying the theorem with a very large relative Reynolds number $1+\\nu^{-1}\\|\\Omega(1)\\|_{L^2(m_0)}$. If an accurate numerical integration of the perturbed vorticity equation shows that $\\sup_{t\\ge T_1} t^{1/2}\\|\\Omega(t)-M(\\Omega(1))G\\|_{L^2(m)}$ grows without bound as the datum size increases, or that a two-vortex initial profile leaves a permanent two-hump residue rather than collapsing to a single Gaussian, then (1.17) and Remark 1.4 would be false.","supporting_citations":[{"cited_title":"Ben-Artzi, Global solutions of two-dimensional Navier-Stokes and Euler equations,Arch","cited_arxiv_id":null,"evidence_quote":"establishes global solvability of the two-dimensional Navier-Stokes vorticity equation with $L^1$ data, used as the basis for Theorem 1.1."},{"cited_title":"Global solutions of two-dimensional Navier-Stokes and Euler equations","cited_arxiv_id":null,"evidence_quote":"supplies the uniqueness argument that fixes the $L^1$ mild solution of the integrated vorticity equation."},{"cited_title":"Gallay and C","cited_arxiv_id":null,"evidence_quote":"provides the model vortex-asymptotics result for 2-D Navier-Stokes that this paper extends to a Couette background."},{"cited_title":"Helffer,Spectral theory and its applications, Cambridge Studies in Advanced Mathematics,139, Cambridge University Press, Cambridge, 2013","cited_arxiv_id":null,"evidence_quote":"supplies the spectral theory of Fokker-Planck operators used for the weighted semigroup estimates."},{"cited_title":"H¨ ormander, Hypoelliptic second order differential equations,Acta Math.,119(1) (1967), 147–171","cited_arxiv_id":null,"evidence_quote":"is the hypoellipticity tool used to gain horizontal derivatives in the energy estimate."},{"cited_title":"Kato, The Navier-Stokes equation for an incompressible fluid inR 2 with a measure as the initial vorticity,Differ","cited_arxiv_id":null,"evidence_quote":"provides the measure-valued well-posedness context and the Nash-Moser iteration estimate used in Section 2."}],"review_version":1}