{"id":"d466f13f-7766-4ab0-89b2-e2b9c137877f","arxiv_id":"2506.03351","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":2,"one_line_summary":"Using boundary-layer asymptotics, the authors claim a fractional Patlak-Keller-Segel equation with a fractional Neumann-type boundary condition, but the derivation has a scaling inconsistency.","lead":"This paper derives, from a kinetic run-and-tumble model of chemotaxis, a fractional diffusion equation and a no-flux boundary condition for cells with reflecting walls. The result is meant to extend classical Patlak-Keller-Segel chemotaxis to fractional (superdiffusive) motion in bounded domains.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Section 3 scaling makes the fractional collision term enter at order ε^{(2α−3)/(α−1)}, which is below the transport order ε for every 1<α<2; the claimed parabolic equation (22) does not follow from the stated balances in (14) and (18).","rationale":"The paper's central claim, Proposition 3.1 and equation (22), depends entirely on the asymptotic expansion in Section 3. The reader's weakest assumption correctly targets the scaling choice γ=1/2, μ=(2−α)/(2(α−1)). I checked the exponents in (14) and (10): the fractional collision term is of order ε^{(2α−3)/(α−1)}, which is larger than the transport term (order ε) and the time-derivative term (order ε²) for all 1<α<2. Consequently, the order at which the flux w1 is determined is not an order where ∂t u0 appears. The paper's equation (18) is also indexed incorrectly for the claimed conclusion: it is written per m, so w1 (from m=1) cannot be inserted into the m=0 equation to produce ∂t u0. Even if one attempts to sum the conservation law over m, the standard ε² balance would be ∂t u0 + c0 n∇·w1=0, but this is not what is derived from (18) as written, and for non-special α the integer-power ansatz is not compatible with the non-integer collision exponents. I found no independent support in the manuscript: no numerics, no machine-checked verification, and no alternative derivation of (24). The concern is an internal ordering inconsistency, not a disagreement with prior consensus. The reader's REJECT verdict remains appropriate.","tokens_in":12259,"tokens_out":26680,"duration_ms":293331,"concrete_test":"Run a symbolic Chapman–Enskog expansion of (14), keeping every term of (10) including the ε^{1−γ}(v·∇) term omitted in (17), for a generic value α=1.8 (so μ=(2−1.8)/(2·0.8)=1/8). Collect the coefficients of ε^{−1/4}, ε^{3/4}, ε^1, and ε^2 in both the velocity-moment equation and the integrated conservation law. Verify whether (i) the ε^1 equation is balanced without forcing ∇u0=0, and (ii) the ε^2 equation reduces to (22). If either fails, Proposition 3.1 is not supported. A control run at α=3/2 (integer collision exponents) shows what a consistent limit should look like.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the ordering fixed after (14): with γ=1/2 and μ=(2−α)/(2(α−1)), the collision terms in (14) have powers ε^{−2μ}=ε^{−(2−α)/(α−1)} and ε^{1−2μ}=ε^{(2α−3)/(α−1)}, while the left-hand side has transport at order ε and time derivative at order ε². For every 1<α<2, (2α−3)/(α−1)<1, so the fractional/chemotactic contribution enters at an order at which no ∂t term is present. Moreover, the ansatz (12)–(13) expands in integer powers of ε, but for generic α the collision exponents are non-integer, so 'taking the ε^m-component' in (14) is not a valid power balance. The conservation equation (18) is written componentwise with the same index m (ε²∂t u_m + ε c0 n∇·w_m=0); when w0=0, the m=0 equation forces ∂t u0=0, while the m=1 equation involves ∂t u1, not ∂t u0. The step 'From (18) we obtain (22)' therefore pairs a time derivative of u0 from m=0 with a flux w1 from m=1 without a consistent ordering. Proposition 3.1 and the boundary condition (24) rest on this invalid balance.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper aims to derive macroscopic boundary conditions for the fractional Patlak–Keller–Segel equation (1) from a kinetic transport model (11) with specular reflection at the boundary. The method is formal matched asymptotics: an interior expansion (12)–(13) is inserted into the scaled kinetic equation, yielding a claimed macroscopic equation (22) for the leading interior density, and a boundary-layer expansion (25) is used to obtain a reflection condition and a final no-flux condition (24). The central claim is Proposition 3.1, which states that the interior density satisfies (22) with the reflective boundary condition (24). The paper also discusses half-space and curved-boundary layers and a solvability condition in Section 6.","tokens_in":12650,"tokens_out":17071,"duration_ms":188882,"significance":"If the derivation were correct, the paper would provide a kinetic justification for fractional Neumann-type boundary conditions in chemotaxis, a topic of current interest in nonlocal models and numerical analysis. The formal use of boundary-layer methods after Alt and Larsen is appropriate, and the limiting case α→2 is informally checked. However, the central asymptotic balance in Section 3 is not valid as written: the scaling chosen makes the fractional and chemotactic collision terms enter at orders larger than the transport and time-derivative terms, so the claimed parabolic equation (22) does not follow from the stated conservation structure. Since the boundary condition (24) is asserted in Proposition 3.1 and is not explicitly derived from the later solvability analysis, the main contribution is not established. The paper therefore does not, in its present form, support its advertised conclusions.","major_comments":[{"comment":"The scaling γ=1/2 and μ=(2−α)/(2(α−1)) makes the two right-hand-side terms in (14) of order ε^{-2μ} and ε^{(2α−3)/(α−1)}. For every 1<α<2, (2α−3)/(α−1)<1 and −2μ<0, so both terms are larger than the transport term of order ε and the time derivative of order ε² in the same equation. Consequently the m=0 conservation equation (18), with w_i0=0, gives ∂t u_i0=0, while the flux w_i1 from the next-order momentum balance enters the flux-divergence term at order ε, not ε². The sentence 'From the conservation equation (18) we obtain (22)' therefore pairs ∂t u_i0 from the m=0 equation with w_i1 from a different order, without a consistent ordering that would justify such a balance. This invalidates Proposition 3.1 and the boundary condition (24), which rest on (22).","section":"§3, Eqs. (14)–(22)"},{"comment":"The 'ε^m-component' extraction in (14) is not a well-defined power balance when the exponents on the right-hand side are non-integer. For example, α=7/5 gives μ=3/4 and hence ε^{-2μ}=ε^{-3/2}, which is not an integer power, while the ansatz (12) expands in integer powers of ε. The paper also switches between ϵ and ε=√ϵ without consistently tracking exponents, and the term −ε^{1−γ}c0/(2−α)(v·∇) appearing in (10) is omitted from (14) and from B_{2μ−1} in (17), so the equation actually expanded is not exactly the one obtained from (10).","section":"§3, Eqs. (12)–(17)"},{"comment":"The reflective boundary condition (24) is stated in Proposition 3.1 without derivation, and the later solvability analysis in Section 6.2 does not explicitly produce (24). Equation (72) gives a condition involving an unknown function h1 and the prompt-reflection operator (1−P), but no argument is supplied that this condition is equivalent to nc0χν·(u_i0∇ρ)+nc0∂ν^{α−1}(Cαu_i0)=0. In addition, the symbol ∂ν^{α−1} in (24) is never defined, despite being central to the claimed boundary condition.","section":"§6.2 and Proposition 3.1"},{"comment":"The proposition claims that the finite sum f_i^{(N)} 'satisfies the equation (11)', but the construction is only a formal asymptotic expansion; no existence proof for the coefficients u_i_m, w_i_m is given, and no error estimate or consistency argument at all orders is provided. The wording overstates what the formal matched-asymptotics computation establishes.","section":"Proposition 3.1"}],"minor_comments":[{"comment":"After integrating (14) over v, the right-hand side vanishes exactly by (6), so (18) contains no information from the collision operator; this makes the subsequent use of (18) to obtain a parabolic equation especially confusing and should be explained if the authors revise.","section":"§3, Eq. (18)"},{"comment":"The definitions of g^b_m and the factors ε^{2μ+2}, ε^{2μ+1}, and ε in (31) should be written in terms of a single small parameter; the current notation mixes ε and ϵ without an explicit dictionary.","section":"§4, Eq. (31)"},{"comment":"The expression Δd (ν·w^b_m) is not clearly defined; if Δd denotes the Laplacian of the distance function, the displayed identity is dimensionally inconsistent and needs a careful derivation in terms of the curvature κ(ξ).","section":"§5, Eq. (52)"}],"recommendation":"reject","confidential_remarks":"The main result Proposition 3.1 and equation (22) are not supported by the asymptotic calculation, and the key scaling is taken from the authors' own earlier paper [7] without resolving the ordering inconsistency. I do not see a local fix that would preserve the stated claims; a corrected asymptotic ordering would require reworking the interior expansion and likely the boundary-layer scaling as well. The paper may be worth a fresh submission after such a reworking, but in its present form it does not meet the standard for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe honest take: this paper addresses a genuinely open problem — what macroscopic boundary condition does a fractional chemotaxis equation inherit from a kinetic transport model with specular reflection? That question deserves attention. The fractional Neumann-type condition (24) is not in the earlier literature, as far as I can tell. And the boundary-layer machinery borrowed from Alt and Larsen is the right toolkit; the neutron-transport analogy (Section 6) is apt and used carefully.\n\nThat's the good news. The bad news is that the central derivation in Section 3 does not work as written. After choosing γ=1/2 and μ=(2−α)/(2(α−1)), the fractional part of the collision term enters at order ε^{(2α−3)/(α−1)}. For every α∈(1,2) that exponent is less than 1, so the term is bigger than the transport term (order ε) and much bigger than the time derivative (order ε²). The leading-order balance in (14) therefore does not produce the parabolic equation (22). Moreover, extracting \"ε^m-components\" in (14) is questionable when the collision exponents are not integers — which they are for generic α. And the conservation equation (18) is set up per m: with w0=0, the m=0 equation gives ∂t u0=0, while the m=1 equation involves ∂t u1. The step 'From (18) we obtain (22)' silently pairs the m=0 time derivative with the m=1 flux. Proposition 3.1 and the boundary condition (24) rest on that pairing. I also notice a term −ε^{1−γ} c0/(2−α)(v·∇) in the T_ε expansion (10) that disappears in the rewriting (14); at γ=1/2 it's order ε, same as transport, so it can't be dropped without comment.\n\nSections 4–6 inherit the problem. The boundary layer equation (30) and the matching are formal, and the linearized transport solvability in Section 6 depends on assumptions (Θ>0 spanning the null space, λ=0 isolated) that are stated but not verified. The conclusion that the interior limit satisfies the Patlak–Keller–Segel equation and (24) is not supported.\n\nWho gets value from this? People who work on kinetic limits or fractional boundary conditions will find the problem statement and the boundary-layer setup instructive. But the derivation as it stands is not a reliable foundation. A serious referee should look at this, mainly to check the ordering; my expectation is that the current version is not citable as a proof. I would not cite it myself.\n\nRecommendation: send it to peer review — the question is important enough — but the review should be clear that the scaling inconsistency is load-bearing. This needs major revision, not a tweak.","headline":"The paper's claimed fractional chemotaxis equation and Neumann condition don't follow from its own asymptotic scaling; the boundary-layer idea is new but the derivation breaks at leading order.","tokens_in":13086,"tokens_out":6929,"would_cite":false,"duration_ms":72376,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35R11","35Q92","92C17","35B40","35C20"],"pacs":[],"model":"deepseek-v4-flash","headline":"A kinetic run-and-tumble model for chemotaxis yields a fractional no-flux boundary condition in the macroscopic limit.","keywords":["fractional diffusion","chemotaxis","kinetic transport model","boundary layer","reflecting boundary condition","fractional Neumann condition","Patlak–Keller–Segel","asymptotic analysis"],"falsifier":"Run a numerical simulation of the kinetic model (11) with specular reflection for a fixed $\\alpha$ in $(1,2)$, say $\\alpha=1.5$, and compare the interior density to the solution of (22) with (24); if the scaling of the fractional term fails to match the time-derivative order in the conservation equation (18), or if the boundary-layer solution does not satisfy (24) at leading order, the claim would be refuted.","tokens_in":12107,"feed_emoji":"🦠","tokens_out":6523,"duration_ms":63787,"temperature":0.7,"pith_summary":"The paper asks what happens to a fractional chemotaxis equation near a reflecting boundary. Starting from a kinetic transport model of cells that run and tumble, with specular reflection at the boundary, it derives a macroscopic fractional Patlak–Keller–Segel equation in the interior and a matching no-flux boundary condition at the boundary. The boundary condition is nonlocal: it involves a fractional normal derivative of the interior density rather than just the normal gradient. In the limit $\\alpha \\to 2$, both the interior equation and the boundary condition reduce to the classical chemotaxis equation with Neumann conditions. The result matters because it gives a first-principles boundary description for nonlocal cell movement.","feed_headline":"No-flux boundary law derived for fractional chemotaxis","feed_subtitle":"A kinetic run-and-tumble model yields a fractional Neumann condition that reduces to the classical one at α=2.","key_machinery":"The argument is carried by the spectral decomposition of the turn operator $T$ and a matched boundary-layer expansion in $\\varepsilon = \\sqrt{\\bar{\\tau}/T}$. The distribution $f$ is written as $\\frac{1}{|S|}(u + n v\\cdot w)$ plus terms orthogonal to linear polynomials; the eigenvalues of $T$ fix the coefficients $C_\\alpha$ and $\\chi$. The scaling $\\gamma = 1/2$, $\\mu = (2-\\alpha)/(2(\\alpha-1))$ is chosen so that the fractional flux enters at the same order as the drift and the time derivative. Near the boundary, the gradient is split into normal and tangential parts in a coordinate $r = \\mathrm{dist}/\\varepsilon^{1/(\\alpha-1)}$, producing a half-space equation whose solvability yields the no-flux condition (24).","core_discovery":"The central claim is Proposition 3.1: for a smooth bounded domain with specular reflection, the leading-order interior density $u_0$ solves $\\partial_t u_0 = n c_0 \\nabla \\cdot \\left(C_\\alpha \\nabla^{\\alpha-1} u_0 - \\chi u_0 \\nabla \\rho\\right)$ subject to $n c_0 \\chi \\, \\nu \\cdot (u_0 \\nabla \\rho) + n c_0 \\, \\partial_\\nu^{\\alpha-1}(C_\\alpha u_0) = 0$ on $\\partial\\Omega$. The derivation uses an $\\varepsilon$-expansion of the kinetic equation, matching an interior solution to a boundary layer that satisfies a half-space transport equation. The boundary condition expresses conservation of particles at the wall through a balance between chemotactic flux and a fractional normal flux; the paper also shows that the classical Patlak–Keller–Segel equation with Neumann boundary conditions is recovered as $\\alpha \\to 2$.","pith_inferences":["If the scaling assumption is stable under perturbations, the same boundary-layer construction should extend to other reflection laws (e.g., diffuse reflection), producing a family of nonlocal boundary conditions parameterized by the reflection kernel $p$.","The derivation suggests a testable prediction: for run-and-tumble organisms with heavy-tailed run lengths, the population flux through a boundary should depend on a fractional normal derivative of the density, so measuring boundary accumulation could estimate $\\alpha$ independently of bulk measurements.","A direct numerical comparison of the kinetic model (11) with (22) plus (24) for intermediate $\\alpha$ would give a quantitative check of the order-of-limits assumption, revealing whether corrections beyond first order are needed near the boundary.","One could also ask whether the same matched expansion produces a fractional Dirichlet condition when the boundary is absorbing; the paper does not do this, but its half-space problem supplies the needed machinery."],"forward_implications":["Proposition 3.1 gives a mathematically explicit macroscopic model: (22) with (24) is the first-order description of the kinetic process, so it can be used as the continuum limit in simulations of fractional chemotaxis.","Mass is conserved at leading order: the matching of interior and boundary-layer solutions, equations (38)–(39), holds with the no-flux condition.","The classical limit is smooth: as $\\alpha \\to 2$, the fractional normal derivative $\\partial_\\nu^{\\alpha-1}$ becomes the usual normal derivative, so the boundary condition reduces to the standard Neumann no-flux condition.","The boundary-layer thickness scales as $\\varepsilon^{1/(\\alpha-1)}$, which grows as $\\alpha \\to 1^+$; nonlocal effects therefore extend further into the domain as the transport becomes more ballistic.","The reflection-operator formalism in Section 6 characterizes admissible boundary data and provides a route to other boundary closures for the same kinetic model."],"supporting_citations":[{"why":"Supplies the boundary-layer and singular perturbation method used for the kinetic model.","marker":"[1]"},{"why":"Provides explicit half-space solution kernels $W$ and $G$ used to represent the boundary-layer solution.","marker":"[4]"},{"why":"Gives the neutron transport theory background for the reflection operator and the half-space boundary value problem.","marker":"[5]"},{"why":"Provides the fractional Patlak–Keller–Segel equation and the scaling choices $\\gamma=1/2$, $\\mu=(2-\\alpha)/(2(\\alpha-1))$ used in the interior expansion.","marker":"[7]"},{"why":"Defines the classical Patlak–Keller–Segel model that must be recovered at $\\alpha=2$.","marker":"[8]"},{"why":"Supplies the asymptotic transport theory and half-space solution formalism used for the boundary layer.","marker":"[10]"}],"fun_headline_variants":["Fractional chemotaxis gets its no-flux boundary law","Boundary layer method derives fractional Neumann condition","Kinetic run-and-tumble model yields fractional flux law","New boundary rule for fractional diffusion in chemotaxis","Macroscopic limit gives fractional no-flux condition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result stands on the scaling choice $\\gamma=1/2$ and $\\mu=(2-\\alpha)/(2(\\alpha-1))$, which makes the fractional flux, the transport term, and the time derivative enter the macroscopic balance at the same order; if that balance is wrong for some $\\alpha$, the derived equation (22) is not the leading-order limit.","fun_headline_variants_meta":{"raw":{"variants":["Fractional chemotaxis gets its no-flux boundary law","Boundary layer method derives fractional Neumann condition","Kinetic run-and-tumble model yields fractional flux law","New boundary rule for fractional diffusion in chemotaxis","Macroscopic limit gives fractional no-flux condition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000316,"raw_usage":{"total_tokens":1731,"prompt_tokens":828,"completion_tokens":903,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":444,"completion_tokens_details":{"reasoning_tokens":827}},"tokens_in":444,"tokens_out":903,"duration_ms":8858,"temperature":1.0,"reasoning_tokens":827,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:06:54.503419+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a numerical simulation of the kinetic model (11) with specular reflection for a fixed $\\alpha$ in $(1,2)$, say $\\alpha=1.5$, and compare the interior density to the solution of (22) with (24); if the scaling of the fractional term fails to match the time-derivative order in the conservation equation (18), or if the boundary-layer solution does not satisfy (24) at leading order, the claim would be refuted.","supporting_citations":[{"cited_title":"Singular perturbation of differential integral equations describing biased random walks.Journal für die reine und angewandte Mathematik, 322:15–41, 1981","cited_arxiv_id":null,"evidence_quote":"Supplies the boundary-layer and singular perturbation method used for the kinetic model."},{"cited_title":"Nuclear reactor theory","cited_arxiv_id":null,"evidence_quote":"Provides explicit half-space solution kernels $W$ and $G$ used to represent the boundary-layer solution."},{"cited_title":"Neutron transport theory","cited_arxiv_id":null,"evidence_quote":"Gives the neutron transport theory background for the reflection operator and the half-space boundary value problem."},{"cited_title":"Fractional Patlak– Keller–Segel equations for chemotactic superdiffusion.SIAM Journal on Applied Math- ematics, 78(2):1155–1173, 2018","cited_arxiv_id":null,"evidence_quote":"Provides the fractional Patlak–Keller–Segel equation and the scaling choices $\\gamma=1/2$, $\\mu=(2-\\alpha)/(2(\\alpha-1))$ used in the interior expansion."},{"cited_title":"Initiation of slime mold aggregation viewed as an instability","cited_arxiv_id":null,"evidence_quote":"Defines the classical Patlak–Keller–Segel model that must be recovered at $\\alpha=2$."},{"cited_title":"Asymptotic theory of the linear transport equation for small mean free paths","cited_arxiv_id":null,"evidence_quote":"Supplies the asymptotic transport theory and half-space solution formalism used for the boundary layer."}],"review_version":1}