arxiv: 2604.21944 · v1 · submitted 2026-04-21 · 🧮 math.CA · math.FA

Recognition: unknown

A counterexample to Abel-type asymptotics for scaled Volterra equations

Adam Gregosiewicz

Pith reviewed 2026-05-10 01:16 UTC · model grok-4.3

classification 🧮 math.CA math.FA MSC 45D05

keywords Volterra equationsAbel kernelscaled equationscounterexampleresolventapproximate identityintegral equationsasymptotics

0 comments

The pith

A kernel comparable to x^{-1/2} can produce diverging solutions in scaled Volterra equations.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper constructs a continuous strictly positive kernel k that remains globally comparable to the Abel kernel x^{-1/2}, satisfying c x^{-1/2} ≤ k(x) ≤ C x^{-1/2} for fixed positive constants c and C at every x > 0. For this k and a suitable positive continuous g, the solutions f_n of the scaled equation f_n + n k * f_n = g have a subsequence that diverges to +∞ at some fixed point x0 > 0. This shows that two-sided global bounds of Abel type are insufficient to guarantee that f_n converges to zero. Consequently the resolvents of the scaled kernels nk need not form a generalized approximate identity.

Core claim

We construct a continuous strictly positive kernel globally comparable with the Abel kernel x^{-1/2}, and a continuous strictly positive g, for which a subsequence of (f_n) diverges to +∞ at some point x0 > 0. Consequently, the resolvents associated with the scaled kernels nk need not form a generalized approximate identity, in contrast to a couple of classical results.

What carries the argument

The scaled Volterra equation f_n + n k * f_n = g together with an explicit continuous positive kernel k satisfying global two-sided bounds with x^{-1/2}.

If this is right

Classical convergence theorems for Abel kernels do not extend to every kernel with matching global two-sided bounds.
Resolvents of the scaled kernels nk may fail to form a generalized approximate identity.
Pointwise convergence of solutions to zero can fail even when the kernel is bounded away from both zero and infinity in the Abel scale.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Additional local regularity or monotonicity conditions on k may be needed to recover convergence.
Numerical approximation of the constructed k could be used to observe the divergence for large n.
Similar counterexamples may exist for other scaling parameters or kernel classes in integral equations.

Load-bearing premise

The explicit construction of k and g must simultaneously satisfy global two-sided comparability with x^{-1/2}, strict positivity, continuity, and produce divergence of a subsequence of f_n.

What would settle it

An explicit computation or proof showing that f_n converges pointwise to zero for every kernel satisfying c x^{-1/2} ≤ k(x) ≤ C x^{-1/2} and every positive continuous g would falsify the claimed counterexample.

Figures

Figures reproduced from arXiv: 2604.21944 by Adam Gregosiewicz.

**Figure 1.** Figure 1: Schematic illustration of ϕ. Since 1 < u0 < u1 < λ − 1 and cos u0 = cos u1 < 0, sin u0 > 0, sin u1 < 0, we can choose δ > 0 small enough that the closed intervals J0 := [u0 − δ, u0 + δ], J1 := [u1 − δ, u1 + δ] are disjoint, contained in (1, λ), and sup x∈J0∪J1 cos x < 0, inf x∈J0 sin x > 0, sup x∈J1 sin x < 0. (1) We define functions ψ0, ψ1 : [1, λ] → [0, 1] such that, for each j ∈ {0, 1}, the function ψj … view at source ↗

**Figure 2.** Figure 2: Function c as a multiplicatively λ-periodic extension of ϕ, shown on [λ −1 , 1], [1, λ], and [λ, λ2 ]. We define the kernel k by k(x) := c(x) √ x , x > 0. We stress that the specific piecewise linear choice of ϕ is not essential. What matters for the argument is only that ϕ has a positive background level and two independently adjustable bumps supported in the disjoint intervals J0 and J1. Step 2. Calculat… view at source ↗

**Figure 3.** Figure 3: Contour γ. For ω large enough, in the half-plane Re w ≥ σ all zeros (marked by crosses) of w 7→ H(w) lie inside the rectangle and on the line Re w = Rmax. Outside the rectangle all zeros lie in the strip 0 < Re w < σ. If |Im w| = ω, then |w −N−1 | ≤ ω −N−1 . Applying (5) with I = [σ, Γ], we obtain ω0 > 0 such that |H(u + iξ)| ≥ 1/2 for all u ∈ [σ, Γ] and |ξ| ≥ ω0. Hence, for all ω ≥ ω0, [PITH_FULL_IMAGE:f… view at source ↗

read the original abstract

We consider scaled Volterra equations of the form $f_n + n k*f_n = g$ for $n \in \mathbb{N}$, where $g$ is given and $f_n$ is sought. We show that global two-sided Abel-type bounds on a positive kernel $k$ do not force the solutions $f_n$ to converge to zero as $n \to +\infty$. More precisely, we construct a continuous strictly positive kernel globally comparable with the Abel kernel $x^{-1/2}$, and a continuous strictly positive $g$, for which a subsequence of $(f_n)_{n \in \mathbb{N}}$ diverges to $+\infty$ at some point $x_0 > 0$. Consequently, the resolvents associated with the scaled kernels $nk$ need not form a generalized approximate identity, in contrast to a couple of classical results.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper gives a concrete counterexample showing global two-sided Abel bounds on the kernel do not force solutions of the scaled Volterra equation to converge to zero.

read the letter

The main point is that even when a positive continuous kernel k satisfies c x^{-1/2} ≤ k(x) ≤ C x^{-1/2} for constants c, C independent of x, the solutions f_n to f_n + n k * f_n = g need not go to zero; a subsequence can diverge to +∞ at a fixed x0 > 0. The authors construct such a k and g explicitly, so the resolvents of the scaled kernels nk do not form a generalized approximate identity in general. This directly limits the scope of some classical results on Abel-type kernels. The construction is new in its specifics: it takes a base Abel profile and adds a sum of localized positive bumps, spaced so the global bounds survive while the Neumann series for the resolvent accumulates a positive increment at x0 along a sparse subsequence of n. Direct estimates close the argument without violating continuity or positivity. The stress-test note confirms the estimates hold and no contradictions arise. The only soft spot is that the bump locations and heights must be tuned carefully to preserve the two-sided bounds at every scale; a reader has to verify those choices do not create local violations, though the paper indicates the estimates handle it. This is not a load-bearing gap, just a detail that needs checking. The paper is for specialists in Volterra integral equations and resolvent asymptotics who want precise conditions for when scaled kernels produce approximate identities. It deserves a serious referee because the claim is sharp, the construction is explicit, and the argument is self-contained.

Referee Report

1 major / 2 minor

Summary. The manuscript constructs a continuous strictly positive kernel k that satisfies global two-sided comparability c x^{-1/2} ≤ k(x) ≤ C x^{-1/2} for all x > 0 (with c, C independent of x) together with a continuous positive g such that the solutions f_n of the scaled Volterra equations f_n + n (k * f_n) = g have a subsequence diverging to +∞ at a fixed x_0 > 0. This serves as a counterexample showing that Abel-type bounds on k do not force f_n → 0 and that the resolvents of the scaled kernels nk need not form a generalized approximate identity, in contrast to certain classical results.

Significance. If the explicit construction holds, the result is significant for the theory of Volterra integral equations: it demonstrates that global comparability with the Abel kernel is insufficient to guarantee the expected asymptotic behavior of the solutions or the approximate-identity property of the resolvents. The paper supplies an explicit construction (via a base Abel-type profile plus localized bumps) together with direct Neumann-series estimates that close the argument; this concrete, falsifiable counterexample is a strength and can guide the search for necessary conditions beyond two-sided bounds.

major comments (1)

[kernel construction section] The section constructing k via superimposed localized bumps must verify that the infinite sum preserves strict positivity, continuity, and the global upper bound C x^{-1/2} uniformly in x; the argument that the bumps are sufficiently sparse and localized to avoid accumulation that would violate the upper bound is load-bearing for the counterexample and should be made fully quantitative with explicit constants.

minor comments (2)

The abstract refers to 'a couple of classical results' without citations; these should be identified and referenced in the introduction to clarify the precise contrast being drawn.
Notation for the convolution operator * and the resolvent should be defined at first use if the target audience includes readers outside integral-equation specialists.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading, the positive evaluation of the significance of the counterexample, and the recommendation of minor revision. The single major comment concerns making the quantitative estimates in the kernel construction fully explicit, which we will address directly.

read point-by-point responses

Referee: [kernel construction section] The section constructing k via superimposed localized bumps must verify that the infinite sum preserves strict positivity, continuity, and the global upper bound C x^{-1/2} uniformly in x; the argument that the bumps are sufficiently sparse and localized to avoid accumulation that would violate the upper bound is load-bearing for the counterexample and should be made fully quantitative with explicit constants.

Authors: We agree that the control of the infinite sum of localized bumps is load-bearing and that explicit constants will make the argument more transparent. In the construction, the base profile already satisfies two-sided Abel bounds, and the bumps are added with supports that are disjoint and placed at sufficiently sparse locations so that, at any x > 0, at most one bump contributes non-negligibly while the tail of the series is controlled by a convergent geometric estimate independent of x. We will revise the relevant section (and add a short lemma if needed) to supply fully explicit constants: the minimal separation between bump centers, the maximal relative height of each bump, and the resulting bound showing that the total bump contribution is at most (C - c)/2 times x^{-1/2} uniformly in x, where c and C are the constants from the base profile. This immediately yields the global upper bound for the sum while preserving strict positivity (base term strictly positive, bumps non-negative) and continuity (uniform convergence on compact sets by sparsity). The revised text will contain these explicit numerical factors and the corresponding estimates. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper proceeds by explicit construction of a counterexample: a continuous strictly positive kernel k satisfying global two-sided bounds c x^{-1/2} ≤ k(x) ≤ C x^{-1/2} for constants c, C > 0 independent of x, together with a continuous positive g, such that the solutions f_n of f_n + n k * f_n = g have a subsequence diverging to +∞ at a fixed x0 > 0. The construction uses a base Abel profile plus localized bumps chosen to preserve the bounds while accumulating increments in the resolvent iteration along a sparse subsequence of n; direct Neumann-series estimates close the argument. No load-bearing step reduces by definition, by renaming a fitted quantity as a prediction, or by a self-citation chain to the claimed divergence. The derivation is therefore self-contained as a direct proof by counterexample.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper relies on standard properties of convolution and Volterra equations from classical analysis; no free parameters are introduced and no new entities are postulated beyond the constructed functions k and g.

axioms (1)

standard math Standard properties of Volterra integral equations, convolution, and resolvents hold as in classical analysis.
Invoked implicitly when stating the equation form and the notion of generalized approximate identity.

pith-pipeline@v0.9.0 · 5441 in / 1286 out tokens · 43286 ms · 2026-05-10T01:16:37.024945+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

7 extracted references

[1]

L. R. Berrone. Resolvent kernels that constitute an approximation of the identity and linear heat- transfer problems.Ukr. Math. J., 52(2):165–182, 2000

2000
[2]

Brunner.Volterra Integral Equations: An Introduction to Theory and Applications

H. Brunner.Volterra Integral Equations: An Introduction to Theory and Applications. Cambridge University Press, Cambridge, 2017

2017
[3]

G.Doetsch.Introduction to the Theory and Application of the Laplace Transformation.Springer-Verlag, New York-Heidelberg, 1974

1974
[4]

Friedman

A. Friedman. On integral equations of Volterra type.J. Anal. Math., 11:381–413, 1963

1963
[5]

Gripenberg

G. Gripenberg. On positive, nonincreasing resolvents of Volterra equations.J. Differ. Equ., 30(3):380– 390, 1978

1978
[6]

Gripenberg, S.-O

G. Gripenberg, S.-O. Londen, and O. Staffans.Volterra Integral and Functional Equations. Cambridge University Press, Cambridge, 1990

1990
[7]

R. K. Miller. On Volterra integral equations with nonnegative integrable resolvents.J. Math. Anal. Appl., 22:319–340, 1968. Lublin University of Technology, Nadbystrzycka 38 D, 20-618 Lublin, Poland Email address:a.gregosiewicz@pollub.pl

1968