{"id":"0884f445-7d8c-43dc-947d-234921ba6cc8","arxiv_id":"1908.01941","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For a broad class of nonlinear parabolic PDEs, adding an incompressible drift with sufficiently small dissipation time ensures global boundedness and exponential decay, including Keller-Segel blow-up prevention and reaction quenching by simple cellular flows.","lead":"This paper proves that adding a fast-swirling fluid motion to certain equations that can blow up in finite time prevents the blow-up, as long as the flow dissipates concentrations quickly. It shows simple cellular flows can be tuned to have this property, keeping Keller-Segel chemotaxis regular and quenching ignition-type combustion.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cellular-flow construction (Theorem 1.3) rests on an uncited uniform heat-kernel lower bound h_{l^2}(x,y) ≥ c l^{-d} in the proof of Theorem 5.3; the bound is plausible but unproved, and Theorem 5.3, Lemma 5.4, and Theorem 5.5 all depend on it.","rationale":"The paper's central theorem 1.2 is clean and self-contained; the energy proofs use only the definition of dissipation time and hypotheses (H1)-(H2). The Keller-Segel application verifies the hypotheses, and the blow-up criterion is cited from [KX16] and reproduced. The soft spot is the construction of concrete cellular flows in Theorem 1.3. There the proof of Theorem 5.3 invokes a strong uniform lower bound on the heat kernel with no proof or reference. This is load-bearing for the examples but not for the abstract suppression theorem. I agree with the reader that this supports CONDITIONAL rather than ACCEPT, and I see no reason to move away from the reader's verdict. The minor typo in Definition 1.1 (L2 norm described as 'non-decreasing' instead of non-increasing) should also be corrected. The acknowledgement that referees pointed out an error in an earlier version adds weight to the need for care around this part of the argument.","tokens_in":21671,"tokens_out":19875,"duration_ms":235609,"concrete_test":"Check the claimed bound for the family of 1/ν-periodic flows by attempting an explicit proof: compute the transition density h_{l^2}(x,y) for u=A∇^⊥ sin(2πx)sin(2πy) on T^2 via Fourier or particle simulation for A=1,10^2,10^4,10^6 and confirm whether min_{x,y} h_{l^2}(x,y) is bounded below by an A-independent constant. More decisively, try to derive h_t(x,y) ≥ c t^{-d/2} e^{-C|x-y|^2/t} for t=l^2 with c,C independent of u using the skew-symmetry of the u·∇ term and Moser iteration; if the constants must depend on ||u||∞ (as they do for general lower-order terms), Theorem 5.3's proof is incomplete and needs either a new argument or a citation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the proof of Theorem 5.3, the authors assert: 'It is well known that there is c>0 such that for any l-periodic flow u, the probability density ... takes values in [c l^{-d}, c^{-1} l^{-d}] when t=l^2.' No proof or citation is supplied. This lower bound is essential: it lets the authors turn the one-point hitting statement from Lemma 5.2 into a statement valid for every starting point x, which is then used to prove Lemma 5.4 and Theorem 5.5. If the bound fails for some family of divergence-free Lipschitz drifts (e.g., if the density can develop holes of size o(l^{-d}) at time l^2 for large-amplitude flows), then Theorem 5.3 is false and the claimed simple cellular flows with arbitrarily small dissipation times in Theorems 1.3, 1.4, and 4.1 are not established by the given argument. The cited [CKRZ08] and [Zla10] lemmas provide upper L1-L∞ bounds for (1.2), not the needed two-sided uniform bound; the standard Aronson-type estimates for operators with drift have constants depending on ||u||∞, so the uniform-in-drift lower bound is not an immediate consequence. A measure-preserving heuristic suggests the bound may be true, but the manuscript does not demonstrate it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the PDE ∂_t θ + u·∇θ = Δθ + N(θ) on the torus, where u is a prescribed divergence-free drift and N obeys two structural hypotheses (H1)-(H2). It introduces the dissipation time τ*(u) of the drift and proves that if τ*(u) is sufficiently small, then the L^2 norm of any mild solution remains uniformly bounded; under an additional small-growth condition (1.4), the norm decays exponentially. The paper then aims to construct simple incompressible flows with arbitrarily small dissipation times by rescaling symmetric cellular flows with large effective diffusivity (Theorem 1.3). These flows are used to prove that blow-up in the Jäger-Luckhaus Keller-Segel system can always be prevented by a suitable ambient drift (Theorem 1.4), and that ignition-type reaction-diffusion equations can always be quenched when the average initial temperature is below the ignition threshold (Theorem 4.1).","tokens_in":21969,"tokens_out":16488,"duration_ms":171626,"significance":"If the results hold as stated, the paper provides a clean and general mechanism for convection-induced singularity suppression: the conditions on the nonlinearity enter only through (H1)-(H2), and the thresholds in Propositions 2.1 and 2.2 are explicit functions of F, G, C0, ε0, with no fitted parameters. The verification of (H1)-(H2) for the Keller-Segel model in Lemma 3.1 is careful and direct, and the exponential decay statement is quantitative. The advertised advantage over prior work is the simplicity of the stabilizing flows: fast cellular flows, which are not mixing, are claimed to have arbitrarily small dissipation times. This is a substantial and attractive claim. However, the proof of the cellular-flow construction rests on an unproved and arguably false uniform heat-kernel lower bound, so the main advertised examples are not established by the submitted version.","major_comments":[{"comment":"The assertion that the probability density h_{l^2}(x,y) of the advection-diffusion process modulo lT^d satisfies h_{l^2}(x,y) ∈ [c l^{-d}, c^{-1} l^{-d}] for every l-periodic divergence-free Lipschitz drift u, with c independent of u and l, is stated without proof or citation. Rescaling to the unit torus shows that this is equivalent to a uniform-in-drift lower bound for the heat kernel of ∂_s - Δ + w·∇ at time 1 for every 1-periodic divergence-free Lipschitz drift w. This is not a consequence of the L^1-L^∞ estimates cited from [CKRZ08] or [Zla10], whose constants may depend on the drift. The claimed uniformity is in fact false for the family w_A = A∇⊥ψ with A→∞: at time 1 the fundamental solution is concentrated near the deterministic flow trajectory and is exponentially small in cells not reached by time 1. Since this lower bound is the only mechanism in the proof that converts the one-point conclusion of Lemma 5.2 into a statement valid for every starting point, and since Lemma 5.4, Theorem 5.5, Theorem 1.3, and the cellular-flow applications in Theorems 1.4 and 4.1 all depend on Theorem 5.3, the construction of flows with arbitrarily small dissipation times is not established by the proof as written.","section":"Section 5, proof of Theorem 5.3"}],"minor_comments":[{"comment":"The phrase \"the L^2-norm of solutions to (1.2) is non-decreasing\" should read \"non-increasing\", since the L^2 norm of an advection-diffusion solution decays in time.","section":"Definition 1.1"},{"comment":"The text says that Theorem 1.3 is proved in Section 6, but the paper has no Section 6; the proof appears in Section 5.","section":"Introduction, last paragraph"},{"comment":"There is a typo with an extra closing parenthesis: ‖θ_{t0+nτ*(u))}‖ should be ‖θ_{t0+nτ*(u)}‖.","section":"Equation (2.1a)"},{"comment":"The choice τ0(y,χ)=τ1(y,y,χ) should explicitly state that y=‖ρ0‖_{L2}; the monotonicity of τ1 then justifies the threshold for arbitrary ρ0, using arρ≤‖ρ0‖_{L2} and ‖ρ0-arρ‖_{L2}≤‖ρ0‖_{L2}.","section":"Proof of Theorem 1.4"},{"comment":"With a = √2Ψ^{-1}(α), the probability P(|√2B_1·e| > a) equals 2α, not 2Cα; the extra C is subsequently absorbed into the final constant, so this is only a notational imprecision.","section":"Proof of Theorem 5.3"},{"comment":"The quantities T0(B) and T1(B) contain integrals of y/F(y), which are not defined if F vanishes on a subinterval of the integration range. Since (H1) only assumes F is increasing and continuous, this case should be addressed explicitly, e.g., by a convention or by treating the case F≡0 on the relevant interval separately.","section":"Propositions 2.1 and 2.2"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the uniform heat-kernel lower bound in Theorem 5.3. On rescaling it appears to be false for large-amplitude cellular flows, so the proof of the paper's headline application is not just missing a citation but likely wrong in the stated form. The general energy estimates and the conditional Keller-Segel/quenching results are sound and could be worth publishing, but the cellular-flow construction needs a substantially new argument or a restriction of the claims. I would ask the editor to enforce a careful revision of Section 5 before reconsidering the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The abstract result in Theorem 1.2 is the real contribution: if a divergence-free drift has dissipation time below a threshold depending only on the nonlinearity and initial data, then L2 norms stay bounded (and decay exponentially under (1.4)). The proof via energy estimates and the inductive use of the dissipation time is clean and correct. I checked Propositions 2.1 and 2.2; they hold together. The verification of (H1)-(H2) for Keller-Segel (Lemma 3.1) is straightforward and correct, and the quenching result (Theorem 4.1) follows from a standard comparison argument plus Proposition 4.2. So the conditional results are fine.\n\nThe problem is in Section 5. The proof of Theorem 5.3 asserts, without proof or reference, that for any l-periodic divergence-free Lipschitz flow u, the heat kernel of the advection-diffusion process at time t = l^2 is bounded below by c l^{-d} uniformly in u. That statement is not 'well known' and, as written, fails: for a shear flow with large amplitude A, the effective diffusivity in the flow direction is O(A^2 l^2), so at time l^2 the density spreads over a distance O(A l^2) in that direction, making the minimum density O(1/(A l^{d+1})) rather than O(l^{-d}). The bound is used to convert the one-point estimate from Lemma 5.2 into a statement uniform in x, and then to prove Lemma 5.4 and Theorem 5.5. If the bound fails, the construction of cellular flows with arbitrarily small dissipation times (Theorem 1.3) is not established. The theorem might still be true for the symmetric cellular flows used in the applications, but the given argument does not cover it.\n\nThere is also a minor typo in Definition 1.1 (the infimum over 't ≥ 0' is written oddly), but that is not serious.\n\nSo my take: the conditional theorems are solid and the paper is worth a serious referee. The gap in the explicit flow construction needs to be fixed, either by giving a correct uniform lower bound for the class of flows actually used or by replacing the argument. Since Theorem 1.4 and the quenching application depend on having explicit flows, the paper should not be accepted in its current form. But the main idea is good and likely repairable. I would send it to peer review with a request for major revision.","headline":"Solid abstract criterion for blow-up suppression via small dissipation time, but the construction of explicit cellular flows rests on an unproved and likely false uniform heat-kernel lower bound.","tokens_in":22538,"tokens_out":4196,"would_cite":true,"duration_ms":44413,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B44","35B27","35Q35","76R05"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that a single quantitative property of a stirring flow—its dissipation time—can be made small enough to prevent blow-up in nonlinear parabolic equations, including Keller-Segel chemotaxis and ignition-type reactions.","keywords":["blow-up suppression","dissipation time","convection-enhanced dissipation","Keller-Segel chemotaxis","reaction quenching","cellular flows","effective diffusivity","nonlinear parabolic PDEs"],"falsifier":"Fix $l$ and take the family of two-dimensional cellular drifts $u_A=A\\nabla^\\perp(\\sin(2\\pi x/l)\\sin(2\\pi y/l))$; a numerical or rigorous computation showing that $\\inf_{x,y}h_{l^2}(x,y)$ can be made smaller than $c l^{-2}$ for a sequence of constants $c\\to0$ by choosing $A$ appropriately would refute the uniform lower bound invoked in Theorem 5.3, and with it the paper's explicit cellular-flow examples in Theorems 1.4 and 4.1.","tokens_in":21433,"feed_emoji":"🌊","tokens_out":9375,"duration_ms":85685,"temperature":0.7,"pith_summary":"The paper establishes a general mechanism by which convection can prevent blow-up in nonlinear parabolic equations. It shows that for any nonlinearity satisfying two structural hypotheses—roughly, that energy production is controlled by a small fraction of the dissipation plus a bounded function of the $L^2$ norm—the $L^2$ norm of every mean-zero solution stays bounded for all time, provided the added divergence-free drift has a dissipation time below a threshold set by the initial data and the nonlinearity. Under an extra small-mass condition the norm decays exponentially. The paper then builds simple stationary cellular flows whose dissipation times can be made arbitrarily small by rescaling, and uses them to prove that chemotactic collapse in the Keller-Segel model can always be prevented, and that ignition-type reactions with sub-threshold average temperature can always be quenched. The interest is that the convection needs no special mixing geometry; only its dissipation time matters.","feed_headline":"Fast-enough stirring suppresses blow-up in nonlinear PDEs","feed_subtitle":"A dissipation-time threshold controls singularity formation in chemotaxis and ignition reactions.","key_machinery":"The load-bearing object is the dissipation time $\\tau_*(u)$, defined as the earliest time by which every mean-zero initial datum is halved in $L^2$ by the advection-diffusion semigroup, uniformly over all starting times. The proof of Theorem 1.2 compares the nonlinear evolution on successive intervals of length $\\tau_*$: an energy estimate using the $1-\\varepsilon_0$ dissipation margin in (H1) shows the norm cannot grow too much on each interval, and the Duhamel formula together with the halving property of $S_{s,s+t}$ prevents the accumulated nonlinear forcing from pushing the norm above $2B+1$. For the flow construction, the key object is the effective diffusivity $D(u)=\\min_e D_e(u)$ of the stochastic process associated with the drift; Theorem 1.3 rescales symmetric cellular flows so that their cells shrink and their effective diffusivity grows, converting large diffusivity into short dissipation time.","core_discovery":"The central discovery is that the growth of solutions to the convection-diffusion-nonlinearity equation $\\partial_t\\theta+u\\cdot\\nabla\\theta=\\Delta\\theta+N(\\theta)$ on the torus is controlled entirely through the dissipation time of the drift. Theorem 1.2 states that if $N$ satisfies hypotheses (H1)--(H2), then for every mean-zero $\\theta_0$ there is a threshold $\\tau_0=\\tau_0(\\|\\theta_0\\|_{L^2},N)$ such that any divergence-free Lipschitz drift with $\\tau_*(u)\\le\\tau_0$ forces the mild solution to obey $\\sup_{t\\in[0,T)}\\|\\theta_t\\|_{L^2}\\le 2\\|\\theta_0\\|_{L^2}+1$; if condition (1.4) holds and $T=\\infty$, the norm decays exponentially. Theorem 1.3 supplies the flows: any sequence of symmetric cellular flows whose effective diffusivity diverges can be rescaled in cell size and amplitude to make $\\tau_*(v_n)\\to0$, with the explicit bound $\\tau_*(v_n)\\le C D(u_n)^{-\\alpha}\\ln(1+D(u_n))$. The Keller-Segel and combustion applications follow by checking (H1)--(H2) for the chemotaxis nonlinearity and by using the $L^1_0\\to L^\\infty_0$ semigroup estimate for quenching.","pith_inferences":["Editorial inference: because the blow-up-prevention condition depends only on $\\tau_*(u)$, one can measure or estimate the dissipation time of a candidate stirrer numerically by evolving the linear advection-diffusion equation, yielding a computable sufficient condition for whether that stirrer will suppress singularities in the nonlinear model.","Editorial inference: the structure of (H1)--(H2) suggests that the same small-dissipation-time criterion should apply to other aggregation or active-scalar models whose nonlinear energy production can be absorbed by a fraction of the dissipation; checking those inequalities for a given model would be a direct extension of Theorem 1.2.","Editorial inference: the quenching proof identifies an explicit quench time $t_0$ set by the reaction rate $\\lambda$ and the margin $\\alpha_0-\\bar\\theta_0$; this could be tested experimentally or numerically by comparing the required stirring strength with predictions from the $L^1_0\\to L^\\infty_0$ estimate."],"forward_implications":["For any nonlinearity obeying (H1)--(H2), adding a divergence-free drift with $\\tau_*(u)\\le\\tau_0$ keeps every mean-zero solution uniformly bounded in $L^2$ on its entire interval of existence, and under condition (1.4) the solution decays exponentially to zero.","Simple time-independent cellular flows, including the standard two-dimensional sine-sine eddy flow, can be rescaled to have arbitrarily small dissipation times in both two and three dimensions.","In the Keller-Segel chemotaxis model on $\\mathbb{T}^2$ and $\\mathbb{T}^3$, every nonnegative smooth initial density can be kept globally regular and driven to the uniform steady state by a suitable steady cellular flow.","For ignition-type reaction-diffusion equations with mean initial temperature below the ignition threshold $\\alpha_0$, a drift with sufficiently small dissipation time always quenches the reaction.","The same proof extends to equations with fractional dissipation $-(-\\Delta)^\\gamma$ in place of $\\Delta$, after adjusting the hypotheses and the energy norm."],"supporting_citations":[{"why":"Supplies the short-time large-probability displacement estimate (Lemma 5.1) from which the density-spreading argument for periodic flows starts.","marker":"[Zla11]"},{"why":"Establishes $D(Au)\\sim A^{1/2}$ for two-dimensional cellular flows, giving the 2D examples in Example 5.7.","marker":"[FP94]"},{"why":"Also establishes the square-root growth of effective diffusivity for random perturbations of 2D Hamiltonian (cellular) flows.","marker":"[Kor04]"},{"why":"Shows the effective diffusivity diverges for the displayed three-dimensional cellular flow family, supplying the 3D examples.","marker":"[RZ07]"},{"why":"Provides the critical-threshold and finite-time blow-up theory for the Keller-Segel system that the new suppression result aims to bypass.","marker":"[JL92]"},{"why":"Supplies the global-regularity criterion (Lemma 3.2) that converts a uniform $L^2$ bound into global existence in the Keller-Segel application.","marker":"[KX16]"},{"why":"Provides the uniform $L^2\\to L^\\infty$ smoothing estimates used in the quenching proof of Proposition 4.2.","marker":"[Zla10]"},{"why":"Gives the qualitative context and a related quenching theorem for relaxation-enhancing flows that Theorem 4.1 makes quantitative and extends to general time-dependent drifts.","marker":"[CKRZ08]"}],"fun_headline_variants":["Fast stirring prevents blow-up in nonlinear PDEs","Dissipation time controls singularity formation","Stirring fast enough tames chemotaxis blow-up","Convection quenches ignition reactions in PDEs","Small dissipation time suppresses PDE singularities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction of the explicit cellular flows rests on a stated-but-not-proved uniform lower bound, $h_{l^2}(x,y)\\ge c l^{-d}$ for the advection-diffusion density on an $l$-torus, asserted to hold for every $l$-periodic divergence-free Lipschitz drift with $c$ independent of the drift; if this bound fails, Lemma 5.4, Theorem 5.5, and the cellular-flow conclusions of Theorems 1.4 and 4.1 are not justified by the given proof.","fun_headline_variants_meta":{"raw":{"variants":["Fast stirring prevents blow-up in nonlinear PDEs","Dissipation time controls singularity formation","Stirring fast enough tames chemotaxis blow-up","Convection quenches ignition reactions in PDEs","Small dissipation time suppresses PDE singularities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000199,"raw_usage":{"total_tokens":1405,"prompt_tokens":1012,"completion_tokens":393,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":323}},"tokens_in":628,"tokens_out":393,"duration_ms":4227,"temperature":1.0,"reasoning_tokens":323,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:01:02.810761+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix $l$ and take the family of two-dimensional cellular drifts $u_A=A\\nabla^\\perp(\\sin(2\\pi x/l)\\sin(2\\pi y/l))$; a numerical or rigorous computation showing that $\\inf_{x,y}h_{l^2}(x,y)$ can be made smaller than $c l^{-2}$ for a sequence of constants $c\\to0$ by choosing $A$ appropriately would refute the uniform lower bound invoked in Theorem 5.3, and with it the paper's explicit cellular-flow examples in Theorems 1.4 and 4.1.","supporting_citations":[],"review_version":1}