{"id":"3d2a0515-77e6-4f18-8748-519a2863c567","arxiv_id":"2607.05158","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Zero-mass solutions of the convection-diffusion equation attain the optimal decay t^{-n/2(1-1/q)-1/2} with self-similar profile given by the adjusted first moment of the heat kernel.","lead":"Global solutions of the convection-diffusion equation with zero-mass initial data decay at the improved rate of the heat kernel's first moment. The paper identifies the exact self-similar profile and the necessary and sufficient moment conditions that make this rate optimal.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The reader correctly identifies both the strongest claim (the sharp asymptotic under zero mass) and the weakest assumption (smallness for the a-priori bound when p is subcritical). The mathematics is classical heat-kernel analysis executed carefully; the self-similar profile A(t) is uniquely determined by the first moments, and the necessity/sufficiency of the moment condition follows at once from the explicit representation of A(1). Because the paper already flags the smallness restriction and the asymptotic theorems are conditional on (2.3), no further load-bearing concern arises. The recommended verdict therefore remains ACCEPT.","tokens_in":18189,"tokens_out":484,"duration_ms":4386,"concrete_test":"Verify the key integral estimate in the proof of (5.1): with the decay (2.3) plugged into ||J_1(t)||_{L^q}, confirm that the resulting time integral converges precisely when p>1+1/(n+1) and yields a remainder o(t^{-n/2(1-1/q)-1/2}). If the exponent calculation fails for any q, the profile argument collapses.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Theorem 2.4) is that, under M_0(u_0)=0 and the a-priori bound (2.3), the solution attains the improved rate t^{-n/2(1-1/q)-1/2} with sharp constant ||A(1)||_{L^q}, and that this constant is positive precisely when a first-moment condition holds. The proofs of Theorems 2.3–2.4 reduce the claim to standard heat-kernel estimates (Lemmas 3.1–3.2) plus the integrability of f(u) that follows from (2.3) once p>1+1/(n+1). The only genuine limitation is the smallness hypothesis needed for (2.3) when p≤1+1/n (Theorem 2.1(2)), but this is already isolated by the reader and does not affect the logical validity of the asymptotic statement under its stated hypotheses. No hidden circularity, unjustified interchange of limits, or gap in the moment analysis appears.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper studies the Cauchy problem for the convection-diffusion equation ∂_t u - Δu = a · ∇f(u) with homogeneous nonlinearity of degree p > 1 and zero-mass initial data M_0(u_0) = 0. Under the weight assumption u_0 ∈ L^1_1 ∩ L^∞, Theorem 2.1 establishes the improved L^q decay ||u(t)||_q ≲ (1+t)^{-n/2(1-1/q)-1/2} for p > 1 + 1/n (no smallness) and for p ≤ 1 + 1/n under a smallness condition in L^1_1 ∩ L^∞. Theorem 2.3 then shows that any solution satisfying the corresponding a-priori bound (2.3) admits the self-similar asymptotic profile A(t) = -(M_1(u_0) - a M_0(ψ_0)) · ∇G_t, with ψ_0 = ∫_0^∞ f(u(\tau)) d\tau. Theorem 2.4 converts this into the sharp statement lim t^{n/2(1-1/q)+1/2} ||u(t)||_q = ||A(1)||_q, and identifies the necessary and sufficient first-moment condition for the constant to be positive. The proofs rely on the Duhamel formula, heat-kernel expansions (Lemmas 3.1–3.2), and previously established higher-order expansions (Lemmas 4.1–4.2).","tokens_in":18452,"tokens_out":1056,"duration_ms":8058,"significance":"The work closes a natural gap left by the classical Escobedo–Zuazua theory and the subsequent higher-order expansions of Zuazua, Kusaba and others: when the mass vanishes, the first-order profiles disappear and the decay improves by exactly one half-derivative. The identification of the critical threshold p > 1 + 1/(n+1) for the Duhamel term to be integrable, the self-similar profile A(t), and the necessary-and-sufficient moment condition for optimality are clean and definitive. The results also supply a positive answer to the open optimality question for the second-order remainder when p = 1 + 3/(2n) under zero mass, and they complete the eta = 1 case left open by Karch–Schonbek. The parallel with the zero-mass Navier–Stokes setting is made explicit and is of independent interest. All estimates are explicit and the logical structure is free of circularity once the cited expansions are granted.","major_comments":[],"minor_comments":[{"comment":"In the statement of Theorem 2.1(2) the smallness threshold ε_0 is asserted to be independent of ||u_0||_{L^1_1 ∩ L^∞}, yet the bootstrap condition (4.1) is written in terms of that norm; a one-line clarification that ε_0 depends only on n, p and a would remove any ambiguity.","section":null},{"comment":"Lemma 4.3 and the subsequent bootstrap are valid for all p > 1 + 1/(n+1), as noted in Remark 4.4; it would be helpful to the reader if this fact were already mentioned in the statement of Theorem 2.1 rather than only in the remark.","section":null},{"comment":"The representation of A(1) used at the end of the proof of Theorem 2.4 (A(1) = (1/2) ∑ (M_{e_j}(u_0) - a_j M_0(ψ_0)) x_j G_1) is elementary but not written earlier; inserting it once when A(t) is introduced would make the positivity argument self-contained.","section":null},{"comment":"A few typographical inconsistencies appear (e.g., the spacing around the multi-index notation in Section 3 and the occasional missing space after commas in the bibliography). They do not affect readability but should be cleaned in the final version.","section":null}],"recommendation":"accept","confidential_remarks":"The manuscript is a solid, technically complete contribution that fits well in a strong analysis journal. The only limitation (smallness for p ≤ 1 + 1/n) is already isolated by the authors and does not undermine the main asymptotic theorems. I see no reason for further delay."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This paper finishes a concrete open case: optimal L^q decay for the convection-diffusion equation when the initial mass vanishes. Under M0(u0)=0 they get the improved rate t^{-n/2(1-1/q)-1/2}, identify the self-similar profile A(t)=-(M1(u0)-a M0(ψ0))·∇Gt, and prove that the rate is attained with positive constant if and only if a first-moment condition holds. That settles the β=1 endpoint left open in Karch–Schonbek (2002) and extends the higher-order expansion program of Escobedo–Zuazua, Carpio, Zuazua and Kusaba to vanishing mass.\n\nWhat works well is the clean reduction. Once the a-priori bound (2.3) is granted, the profile theorem and the optimality criterion follow from standard heat-kernel expansions (Lemmas 3.1–3.2) plus integrability of f(u) that is forced by p>1+1/(n+1). The proofs of Theorems 2.3–2.4 are short, explicit and free of circularity; the moment condition is an independent algebraic statement about the data and the integrated nonlinearity. The upper bound for p>1+1/n is unconditional (via the already-available higher-order expansions), and the critical exponent appears naturally from the Duhamel term.\n\nThe soft spot is real but clearly flagged: when p≤1+1/n the improved decay (and therefore the a-priori bound needed for the profile) requires smallness of u0 in L1_1 igcap L^\tau. Without smallness the paper does not claim the rate. That is a genuine restriction, not a hidden gap, and it does not undermine the asymptotic statement under its stated hypotheses. Everything else—citations, self-citations to earlier expansions, absence of free parameters—looks solid.\n\nThis is for people who already work on large-time asymptotics of scalar parabolic equations or who want a model problem that isolates quadratic convection under a mass constraint. It is not a paradigm shift, but it is a clean, usable completion of an established line. I would send it to referees without hesitation; the mathematics is classical and fully written out. Worth citing if you need the zero-mass rate or the sharp moment criterion.","headline":"Solid, complete resolution of the zero-mass decay problem for scalar convection-diffusion, including the eta=1 case left open by Karch–Schonbek; the only real limitation is the smallness restriction for p ≤ 1+1/n.","tokens_in":19050,"tokens_out":622,"would_cite":true,"duration_ms":5473,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35K58","35B40","35C20"],"pacs":[],"model":"grok-4.5","headline":"Zero-mass data for the convection-diffusion equation force a faster optimal decay rate and a unique self-similar profile built from first moments.","keywords":["convection-diffusion equation","optimal decay rate","zero-mass initial data","asymptotic profile","self-similarity","higher-order expansions"],"falsifier":"Exhibit a zero-mass initial datum (even a small one) for which the first-moment vector $M_1(u_0)-a M_0(\\psi_0)$ vanishes while the solution still decays exactly like $t^{-\\frac{n}{2}(1-\\frac{1}{q})-\\frac{1}{2}}$ in some $L^q$, or construct a large-data solution for $p\\le 1+\\frac{1}{n}$ that decays slower than the claimed rate.","tokens_in":19081,"feed_emoji":"📉","tokens_out":722,"duration_ms":5593,"temperature":0.7,"texified_at":"2026-08-05T21:15:55.405138+00:00","pith_summary":"The paper studies the Cauchy problem for the scalar convection-diffusion equation when the initial mass vanishes. Under that zero-mass condition the usual heat-kernel decay is improved by an extra half power of time. The authors first prove the corresponding upper bound (with a smallness restriction when the nonlinearity is not supercritical) and then construct a self-similar asymptotic profile that is a linear combination of gradients of the heat kernel whose coefficients are first moments of the data and of the integrated nonlinearity. Because the profile is self-similar, the improved decay rate is attained if and only if at least one of those moments is nonzero, and is therefore optimal. The result also closes an earlier gap for the borderline exponent that appears in second-order expansions when mass is zero.","texify_model":"deepseek-v4-flash","texify_usage":{"total_tokens":5750,"prompt_tokens":628,"completion_tokens":5122,"prompt_tokens_details":{"cached_tokens":0},"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":4531}},"feed_headline":"Zero mass forces a faster, optimal decay for convection-diffusion","feed_subtitle":"A unique self-similar profile built from first moments shows the half-power gain is sharp.","key_machinery":"The second-order free-solution expansion (Lemma 3.2) together with a Duhamel remainder that is shown to be $o\\left(t^{-\\frac{n}{2}(1-\\frac{1}{q})-\\frac{1}{2}}\\right)$ once the zero-mass assumption and the a-priori bound are used; the resulting profile $A(t)$ inherits the exact scaling of a first-order heat-kernel derivative.","core_discovery":"Under the zero-mass condition and an a-priori decay bound of the same strength, every global solution satisfies $\\lim t^{\\frac{n}{2}(1-\\frac{1}{q})+\\frac{1}{2}} \\|u(t)\\|_q = \\|A(1)\\|_q$, where $A(t) = -(M_1(u_0)-a M_0(\\psi_0))\\cdot \\nabla G_t$ is the unique self-similar profile of that order; the limit is positive precisely when one of the first-moment conditions $M_{e_j}(u_0)-a_j M_0(\\psi_0)$ is nonzero.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Zero-mass data forces optimal half-power faster decay","Self-similar first-moment profile makes decay rate sharp","First moments fix necessary and sufficient optimal decay","Zero mass yields unique asymptotic profile of order t^{-1/2}","Improved decay holds precisely when first moments nonzero"],"cache_read_input_tokens":128,"weakest_assumption_plain":"When the power $p$ is at most $1+\\frac{1}{n}$ the initial datum must be sufficiently small in the weighted $L^1-L^\\infty$ norm for the improved upper bound to be proved.","fun_headline_variants_meta":{"raw":{"variants":["Zero-mass data forces optimal half-power faster decay","Self-similar first-moment profile makes decay rate sharp","First moments fix necessary and sufficient optimal decay","Zero mass yields unique asymptotic profile of order t^{-1/2}","Improved decay holds precisely when first moments nonzero"]},"model":"grok-4.5","effort":"low","cost_usd":0.006918,"raw_usage":{"total_tokens":1595,"prompt_tokens":663,"num_sources_used":0,"completion_tokens":79,"cost_in_usd_ticks":69180000,"prompt_tokens_details":{"text_tokens":663,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":853,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":663,"tokens_out":79,"duration_ms":6288,"temperature":1.0,"reasoning_tokens":853,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-11T07:53:17.243774+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Exhibit a zero-mass initial datum (even a small one) for which the first-moment vector $M_1(u_0)-a M_0(\\psi_0)$ vanishes while the solution still decays exactly like $t^{-\\frac{n}{2}(1-\\frac{1}{q})-\\frac{1}{2}}$ in some $L^q$, or construct a large-data solution for $p\\le 1+\\frac{1}{n}$ that decays slower than the claimed rate.","supporting_citations":[],"review_version":1}