REVIEW 2 major objections 4 minor 1 cited by
Variable-Radius Disk Transforms and an Area-Integral Problem of Zalcman
T0 review · 2 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash
Pith's one-line read For bounded continuous functions on the unit disk, zero integrals over every variable-radius disk force the function to vanish identically—except for an infinite-dimensional family of smooth unbounded exceptions at subcritical scales.
desk verdict Solid resolution of a named open problem, apart from a printed corollary that contradicts the paper's own theorem. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The proof has two main engines. At the endpoint α=1, a Cayley transform straightens the tangent disks into horodisks, and a horizontal Fourier transform turns the family of area integrals into a generalized Abel-type equation with nonvanishing diagonal, whose injectivity follows by reduction to a second-kind Volterra equation and Gronwall back-propagation. For 0<α<1, the angular modes of normalized disk means satisfy the Euler–Poisson–Darboux equation; hyperbolic coordinates turn the oblique vanishing set into a Dirichlet boundary condition, and conjugation by e^{−3τ/2} produces a wave equation on a finite strip with exponentially decaying coefficients. A coercive energy estimate converts th
What would settle it
A concrete check is to solve the first-mode Volterra equation numerically for a bounded, compactly supported ψ and ask whether Jψ = 0 on (α,1) forces ψ ≡ 0; the theorem predicts no bounded nonzero solution exists. More directly, the decay estimate E(τ) ≤ C e^{−3τ} can be tested for data near the boundary: a slower-than-exponential decay would invalidate the Gronwall back-propagation and would indicate a flaw in the subcritical uniqueness proof.
Extended reading notes
Core claim
The central claim is a complete dichotomy for the transform T_α. For 0<α<1, T_α is injective on bounded continuous functions; for α=1, T_1 is injective on integrable functions; and for every 0<α<1, the kernel inside the space of smooth functions is infinite-dimensional, generated smoothly by compactly supported data and consisting only of functions unbounded near the boundary. The first two statements are uniqueness results: a function whose variable-radius disk means all vanish must itself vanish. The third shows that dropping the boundedness hypothesis destroys uniqueness, so the L∞ condition in the subcritical case is exactly the threshold separating rigidity from non-rigidity.
Load-bearing premise
The proof that bounded solutions are trivial depends on the global boundedness of the function, which is converted into terminal exponential decay; without that boundedness, smooth counterexamples exist, and the authors also flag that the original problem's integral is being read as an area integral rather than a general differential.
Editorial extensions
If this is right
- If f is continuous on the closed disk, bounded, and all variable-radius disk integrals vanish for some 0<α<1, then f is identically zero; in fact every angular Fourier coefficient vanishes in a boundary annulus.
- At α=1 the same uniqueness holds for merely integrable functions, with the conclusion that f = 0 almost everywhere.
- For every 0<α<1 there exist nonzero smooth solutions, and any such solution is necessarily unbounded near the boundary—so continuity alone cannot force rigidity.
- The three statements together give a complete answer to the original problem: under area measure, the answer is 'yes' for bounded continuous functions and 'no' for merely continuous interior functions.
- The method of converting boundedness into exponential decay of a boundary energy is a template that should apply to other rotation-covariant variable-scale problems.
Reading between the lines
- The energy-decay step suggests that boundedness is the sharp condition for uniqueness in any family of rotation-covariant variable-scale disks, not just the boundary-proportional scaling used here; analogous unbounded kernels can be expected when the scaling rule differs.
- Since the smooth kernel is injectively parameterized by C_c^∞((0,α)), one can study the 'degree of non-uniqueness' as α varies: the support of the boundary data controls the size of the constructed null space, and the map is linear, so the kernel carries a natural Frechet-space structure.
- The generalized Abel factorization near the maximal radius of support may extend to arbitrary dimension n: the Euler–Poisson–Darboux equation for spherical means still holds, and the same strip-energy argument should prove injectivity on bounded functions in R^n with analogous unbounded smooth kernels.
- The paper's reading of the printed differential as planar Lebesgue measure is interpretive, not forced; if the original problem intended a complex line integral, the dichotomy proven here would not directly apply, leaving that version open.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This paper studies the variable-radius disk transform T_alpha f(z) = integral over B(z, alpha(1-|z|)) of f(zeta) dA(zeta) for 0<alpha<=1 on the unit disk. The main results are: for 0<alpha<1, T_alpha is injective on C(D) intersect L^infinity(D); at alpha=1, T_1 is injective on L^1(D); and for each 0<alpha<1 there is an injective linear map from C_c^infinity((0,alpha)) into the smooth kernel of T_alpha, with every nonzero element of its image unbounded near the boundary. The proof combines generalized Abel integral equations, a Cayley-Fourier reduction at the endpoint, a hyperbolic-coordinate Euler-Poisson-Darboux energy estimate, and Volterra continuation for the first angular mode. The authors present this as a complete answer to Hayman-Lingham Problem 7.29 under the area-measure interpretation of the printed differential d(zeta).
Significance. Assuming the main theorems are correct, this is a substantial and sharp result. It resolves an open problem in integral geometry under the stated interpretation and, importantly, identifies the exact unboundedness threshold: boundedness is not a technical convenience but the precise uniqueness condition. The paper is essentially self-contained: the Abel lemmas are proved in Section 2, the disk-mean Darboux equation is derived by mollification in Lemma 4.1, the central hyperbolic-coordinate identity and Abel factorizations are explicit, and the energy argument is standard but carefully applied. There are no free parameters or fitted constants. The methods, particularly the coercive strip estimate converting boundedness into boundary decay and the canonical flat Volterra continuation, are original and well matched to the geometry. The main caveat is that the interpretation of the printed differential d(zeta) as planar Lebesgue measure is an explicit premise; under that premise, the results are significant.
major comments (2)
- [Section 1, Corollary 1.3] As printed, Corollary 1.3 is false and contradicts the paper's own Theorem 1.2(iii). The corollary states that if f in C(D) and T_alpha f = 0 then f identically 0 for every 0<alpha<=1. But for 0<alpha<1, Theorem 1.2(iii) gives an injective linear map from C_c^infinity((0,alpha)) into {f in C^infinity(D) : T_alpha f = 0}, and every nonzero element of its image is a nonzero C^infinity(D) function, hence in C(D). The proof of Theorem 1.2(i) uses L^infinity essentially: the constant M enters the gradient bounds (4.9) and the terminal decay E(tau) <= C M^2 e^{-3 tau} (4.12), and the Gronwall argument collapses without it; Section 5 shows the conclusion is false without boundedness. If C(D) in the corollary was intended as C of the closed disk, that assumption must be stated explicitly, since C of the closed disk is contained in L^infinity(D). As written, the announced resolution of Problem 1.
- [Section 1 and Section 6] The manuscript claims a complete answer to Problem 1.1, but the actual answer to the original question 'Must f be holomorphic in D?' is not stated correctly because of the overclaim in Corollary 1.3. The correct resolution under the area-measure reading is: for 0<alpha<1, the answer is negative in general because nonzero smooth unbounded solutions exist, and positive under the additional L^infinity assumption (or if f is continuous on the closed disk); for alpha=1, the answer is positive for f in L^1(D). This dichotomy, not the statement that f identically 0 for every 0<alpha<=1, is the actual content of Theorem 1.2. The abstract, introduction, and Section 6 should state this uniformly so that the central claim is internally consistent.
minor comments (4)
- [Section 1, Problem 1.1] The displayed problem writes dA(zeta), while the surrounding text says the original printed differential is d(zeta). Since the area-measure reading is a premise of the main claim, the display should clearly distinguish the original statement from the interpreted area-measure version.
- [Section 3, after (3.1)] The phrase with the usual continuous interpretation at xi=0 is terse. For clarity, spell out the limiting kernel at xi=0 so that the application of Lemma 2.1 with Q_0(X,y)=2 sinc(0)=2 is unambiguous.
- [Section 5.2, global linearity] The argument that finitely many initial data can be handled by common cutoffs and partitions is compressed. Because the local extension is canonical via uniqueness, the conclusion is credible, but a short formal statement of linearity on finite-dimensional subspaces would improve readability.
- [Section 4.4 and equation (5.4)] Minor notation and typesetting issues: the tilde notation for F_n appears as eFn in some passages, and the square root expression in (5.4) should be typeset unambiguously so the positivity claim is immediate.
Circularity Check
No circularity: all Abel, Darboux, and energy ingredients are proved or cited to external standard sources; no self-citation, fitting, or definitional collapse.
full rationale
The paper's derivation chain is self-contained. All load-bearing analytic tools are either proved in the text (Lemma 2.1, Lemma 2.2, Lemma 4.1, Lemma 4.2, Proposition 3.1, Proposition 5.1) or cited to standard external references (Dautray–Lions for the energy lemma, Nguyen for the Darboux equation, Atkinson and Gorenflo–Vessella for Abel theory). There is no calibration against data, no fitted parameter renamed as a prediction, and no self-citation: the reference list contains no work by Guo or He. The transform T_alpha is defined independently, the uniqueness theorems are proved from the stated hypotheses (L∞ or L1) via differential identities, coercive energy estimates, and Volterra continuation, and the counterexample family is constructed explicitly by solving a first-mode Volterra equation. The fact that the paper's own Corollary 1.3 appears to contradict Theorem 1.2(iii) is an internal correctness/consistency issue, not a circularity of the derivation. Thus the circularity score is 0.
Assumptions & free parameters
assumptions (6)
- standard math Classical Abel operator injectivity (u ↦ ∫₀ˣ u(t)/√(x−t) dt is injective on L¹) and the identity A₀(A₀W)(x) = π∫ₐˣ W(t)dt.
- standard math Disk means (Mh)(z,t) satisfy the Euler–Poisson–Darboux equation u_tt + (3/t)u_t = Δu in the distributional sense for bounded L^∞ data, with gradient bound |∇U| + |U_t| ≤ C t^{-1}‖h‖∞.
- standard math Energy identity E'(τ) = Re(G(τ), Z_τ(τ))_H for second-order equations Z_ττ + PZ = G on the Gelfand triple X ↪ H ↪ X*.
- standard math Coercivity and Poincaré-type inequality q[φ] ≥ (π²/η₀²)‖φ‖²_H for the weighted form q on (0,η₀) with weight w = sinh³η.
- domain assumption Möbius maps send the family {B(z,1−|z|)} of internally tangent disks onto all horodisks of the upper half-plane.
- standard math Approximate-identity convergence of normalized disk means at continuity points of L^∞ data: U(z,t) → h(z) as t↓0.
Cite this review
Pith. "Pith review of Variable-Radius Disk Transforms and an Area-Integral Problem of Zalcman." pith.science (2026). https://pith.science/paper/VMDTF77F
@misc{pith2026260802546,
author = {Pith},
title = {Pith review of: Variable-Radius Disk Transforms and an Area-Integral Problem of Zalcman},
year = {2026},
howpublished = {\url{https://pith.science/paper/VMDTF77F}},
note = {Machine review of arXiv:2608.02546}
}
abstract
For $0<\alpha\leq1$, define $(\mathcal T_\alpha f)(z) :=\int_{B(z,\alpha(1-|z|))}f(\zeta)\,dA(\zeta)$ for $z\in\mathbb D$, where $dA$ is planar Lebesgue measure. We prove that $\mathcal T_\alpha$ is injective on $C(\mathbb D)\cap L^\infty(\mathbb D)$ for $0<\alpha<1$, and that $\mathcal T_1$ is injective on $L^1(\mathbb D)$. In contrast, for each $0<\alpha<1$ there is an injective linear map from $C_c^\infty((0,\alpha))$ into the kernel of $\mathcal T_\alpha$ on $C^\infty(\mathbb D)$; every nonzero function in its image is necessarily unbounded near $\partial\mathbb D$. Under the area-measure interpretation, these results give a complete answer to Hayman--Lingham Problem~7.29, attributed there to L.~Zalcman. The proof combines generalized Abel equations, an Euler--Poisson--Darboux energy argument, and Volterra continuation.
Forward citations
Cited by 1 Pith paper
-
An Infinitesimal Circular Morera Theorem
A continuous function whose circular integrals around every point are o(r^2) must be holomorphic.
Reviewed August 4, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.