Nonlocal gradient, the nonlocal Laplacian and maximum principles

Guillermo Garc\'ia-S\'aez; Jos\'e Camilo Rueda-Ni\~no; Jos\'e Carlos Bellido

REVIEW 3 major objections 3 minor 51 references

Nonlocal gradient, the nonlocal Laplacian and maximum principles

T0 review · 3 major / 3 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The nonlocal ρ-Laplacian built from a radial kernel is, under matching fractional orders, an integro-differential elliptic operator of order 2s, and it satisfies maximum principles under even weaker conditions.

desk verdict The kernel representation for Δρ is genuinely useful, but the main bridge theorem is false as stated because the symbol comparison only holds for large frequencies and fails for compactly supported kernels, including the paper's own truncated Riesz example. read the letter →

arxiv 2607.13161 v1 pith:LO2C6SHA submitted 2026-07-14 math.AP

José Carlos Bellido , Guillermo García-Sáez , José Camilo Rueda-Niño This is my paper · ORCID

classification math.AP MSC 35B5035B5135K0947G20

keywords nonlocalgradientLaplacianintegro-differentialoperatorsmaximumprinciplesconvolutionkernelFouriersymbolfractionalSobolevspaces

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper studies the nonlocal ρ-Laplacian, formed by composing the nonlocal gradient and divergence with a general radial kernel ρ, and proves that it is a convolution operator in disguise: Δ_ρ u = K_ρ * u with an explicit kernel K_ρ. Its main structural result shows that when ρ satisfies natural scale conditions (H0)–(H4) with a single fractional order s = t, the kernel K_ρ belongs to the standard class of integro-differential elliptic operators whose Fourier symbol is comparable to |ξ|^{2s}. This makes the full regularity theory for such operators available to variational models built on nonlocal gradients. The paper also proves strong and weak maximum principles for Δ_ρ that require only the minimal hypothesis (H0) plus nonnegativity of K_ρ, so they cover operators outside the elliptic class. The equivalence of the associated nonlocal Sobolev and Gagliardo spaces is a supporting bridge.

What carries the argument

The load-bearing object is the auxiliary kernel Q_ρ(x) = ∫_{|x|}^∞ ρ(r)/r dr, whose Fourier transform Q̂_ρ(ξ) is proportional to the symbol of the nonlocal gradient. The ρ-Laplacian is built from Q_ρ by K_ρ = −DQ_ρ ∗ DQ_ρ, so its symbol is −4π^2|ξ|^2 Q̂_ρ(ξ)^2. The hypotheses (H3)–(H4) control Q̂_ρ between |ξ|^{s−1} and |ξ|^{t−1} for large |ξ|; when s = t this yields the comparability to |ξ|^{2s} that characterizes the integro-differential class. The kernel K_ρ itself is radial and nonnegative when (4.6) holds, which is what the maximum principles use.

What would settle it

Take the truncated Riesz kernel ρ_δ^s(x) = w_δ(x)/|x|^{n+s−1} with compact support, as in the paper. Since Q_ρ is integrable, its Fourier transform Q̂_ρ is continuous and tends to a finite positive value as ξ→0; hence m_{K_ρ}(ξ) = 4π^2|ξ|^2 Q̂_ρ(ξ)^2 is bounded between c|ξ|^2 and C|ξ|^2 near the origin. For 0 < s < 1, this cannot satisfy λ|ξ|^{2s} ≤ m_{K_ρ}(ξ) for small ξ, directly contradicting the claimed membership K_ρ ∈ K_s(λ,Λ).

Watch

Extended reading notes

Core claim

The central discovery is that the ρ-Laplacian can be written as a single convolution with a radial kernel K_ρ = −DQ_ρ ∗ DQ_ρ, where Q_ρ is the integral of the tail of ρ; its Fourier symbol is m_{K_ρ}(ξ) = −4π^2|ξ|^2 Q̂_ρ(ξ)^2. Using estimates on Q̂_ρ for kernels whose behavior near zero is trapped between two fractional powers (H3)(H4), the paper proves (Theorem 4.18) that if the two powers coincide, s = t, then K_ρ lies in the class K_s(λ,Λ) of symmetric integro-differential operators with λ|ξ|^{2s} ≤ m_K(ξ) ≤ Λ|ξ|^{2s}. This is the bridge from nonlocal-gradient calculus to the modern theory of integro-differential equations. Independently, Proposition 5.2 shows that if K_ρ ≥ 0, then Lu ≥ 0

Load-bearing premise

The theorem's conclusion assumes the symbol compares to |ξ|^{2s} at all frequencies, while the proof only verifies this at large frequencies; small-frequency behavior can differ (e.g., order |ξ|^2 for compactly supported kernels).

Editorial extensions

If this is right

Any regularity result for integro-differential operators of order 2s (e.g., Hölder and Schauder estimates) automatically applies to minimizers of energies built with nonlocal gradients.
The integration-by-parts identity ⟨D_ρ u, D_ρ v⟩ = ⟨u, v⟩_ρ gives a variational formulation of the elliptic equation in a natural way, connecting to Lax–Milgram and existence theory.
The strong maximum principle holds for a broader family of ρ than those in K_s(λ,Λ), since it only relies on nonnegativity of the convolution kernel K_ρ.
The equivalence W^{ρ,2} = H^{ρ,2} shows that nonlocal Gagliardo spaces coincide with potential spaces, making Fourier and functional-analytic tools interchangeable in this setting.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The small-frequency behavior of the symbol is never verified in the paper; for compactly supported kernels the symbol behaves like |ξ|^2 near zero, so the membership theorem as stated appears to need an added large-scale condition or a restricted notion of the class.
One could test the boundaries of the maximum principles by constructing kernels with sign-changing K_ρ or with zero set, to see if the nonnegativity condition (4.6) is necessary.
The potential-theoretic characterization hints at a natural extension to p ≠ 2 via multipliers, possibly defining new nonlocal Besov-type spaces; the paper mentions this as future work.
If the dictionary between nonlocal gradient calculus and integro-differential operators is as strong as the paper suggests, one could transfer nonlinear regularity theory (e.g., for the fractional p-Laplacian) to general kernel settings or use known integro-differential tools for obstacle problems.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: the convolution representation of Δρ is a real and useful observation, but the advertised bridge to Fernández-Real/Ros-Oton does not survive contact with low frequencies. Theorem 4.18 is false as stated, and the maximum-principle section has gaps. The paper deserves a referee, but it needs substantial repair before the central claims can be trusted.

What is actually new and good: Proposition 4.2 identifies Δρu as convolution with Kρ = −DQρ ∗ DQρ, with symbol −4π²|ξ|² Q̂ρ(ξ)². That is a clean, useful observation, and the paper exploits it well to connect the nonlocal gradient calculus to the integro-differential framework. The potential characterization of H^{ρ,2} and the equivalence with the Gagliardo-type space W^{ρ,2} are also solid and worth keeping. The overall program — translating between nonlocal gradient variational theory and integro-differential elliptic theory — is sensible and worth pursuing.

The soft spot is not minor. Theorem 4.18 claims Kρ ∈ K_s(λ,Λ) under (H0)–(H4) with s=t. The proof only establishes symbol bounds for |ξ| ≥ 1/ε, via the estimates on Q̂ρ. But membership in K_s requires comparability λ|ξ|^{2s} ≤ m_{Kρ}(ξ) ≤ Λ|ξ|^{2s} for every ξ ∈ Rⁿ. For compactly supported ρ, Qρ ∈ L¹, so Q̂ρ is continuous and Q̂ρ(0) = ∫Qρ > 0. Hence m_{Kρ}(ξ) = 4π²|ξ|²Q̂ρ(ξ)² behaves like c|ξ|² near ξ = 0, which is not comparable to |ξ|^{2s} for s < 1. The truncated Riesz kernel ρ_δ^s = wδ/|x|^{n+s−1} is compactly supported and satisfies (H0)–(H4) with s=t, so it directly falsifies the theorem as stated. This is not a technicality; it is the central claim of the paper. The theorem could potentially be salvaged by adding a heavy-tail condition, or by proving a weaker statement with a frequency-dependent comparability, but that is not what is written. The abstract's claim of 'minimal assumptions' is also overstated: the maximum principles require (4.6), the nonnegativity of Kρ, which is neither verified nor characterized.

The maximum-principle proofs have gaps. In Proposition 5.2, the iteration step finds a point x₁ = x₀ + h₀ with the same minimum value, but the argument then assumes that repeating this with the same h₀ eventually exits Ω. Nothing guarantees that h₀ points outward, and arbitrary choices of h₀ could cycle or stay inside a bounded domain. A connectivity or chain-of-balls argument over the support of Kρ is needed; it is not supplied. In Proposition 5.4, the derivation of ⟨u⁻,u⁻⟩_ρ = 0 is fine, but concluding u⁻ ≡ 0 requires a Poincaré-type or connectivity fact for H^{ρ,2}_0(Ω) that is not stated. These look repairable rather than hopeless.

Who is this for: researchers in nonlocal gradients, variational calculus, and integro-differential equations. The representation formula alone is worth knowing, but the advertised connection should not be cited as established. A serious referee should see it, with instructions to focus on low-frequency symbol comparability and the maximum-principle rigor. Expect major revision.

Referee Report

3 major / 3 minor

Summary. The paper introduces the nonlocal rho-Laplacian Δ_ρu = div_ρ(D_ρu) associated with a radial kernel ρ, derives a convolution/Fourier representation through the auxiliary kernel Q_ρ, and defines associated Gagliardo-type spaces W^{ρ,2}. The main theoretical claim is that, under hypotheses (H0)–(H4) with s=t, the operator belongs to the integro-differential class K_s(λ,Λ) of Fernández-Real and Ros-Oton. The paper also proves strong and weak maximum/comparison principles for the ρ-Laplacian under only (H0) plus the pointwise positivity of K_ρ.

Significance. The idea of connecting the nonlocal gradient calculus to the regularity theory of integro-differential operators is natural and potentially useful. The Plancherel-based identification of W^{ρ,2} with the potential space H^{ρ,2}, the integration-by-parts formula, and the convolution representation of Δ_ρ are clean and will be useful if the framework is corrected. However, the central membership theorem is false for the compactly supported kernels that the paper itself highlights, and the maximum-principle proofs have important gaps. The stress-test concern is confirmed by the manuscript's own estimates.

major comments (3)

[§4.2, Theorem 4.18] Theorem 4.18 asserts K_ρ ∈ K_s(λ,Λ) under (H0)–(H4) with s=t. Membership requires the global symbol comparability (3.5): 0 < λ|ξ|^{2s} ≤ m_{K_ρ}(ξ) ≤ Λ|ξ|^{2s} for every ξ. The only estimate proved, Corollary 4.17, is restricted to |ξ| ≥ 1/ε. For a compactly supported ρ, such as the truncated Riesz kernel ρ_δ^s = w_δ/|x|^{n+s-1} highlighted in §3.1, Q_ρ has compact support and lies in L¹, so Q̂_ρ is continuous and Q̂_ρ(0)=∫Q_ρ>0. Since m_{K_ρ}(ξ)=4π²|ξ|²Q̂_ρ(ξ)², we get m_{K_ρ}(ξ) ∼ c|ξ|² as ξ→0. For s<1 this is not comparable with |ξ|^{2s}; the lower inequality in (3.5) fails for small ξ. The example satisfies the stated hypotheses, so Theorem 4.18 is false as written. A low-frequency/heavy-tail condition, e.g. Q̂_ρ(ξ) ∼ |ξ|^{s-1} near 0, is needed and is absent.
[§5.1, Proposition 5.2] The proof of the strong maximum principle is not justified. After showing Lu(x_0)=0 at a global minimum, the authors find one h_0 ∈ B_r(a) with u(x_0+h_0)=m, then iterate to obtain x_{k+1} ∈ x_k + B_r(a) with u(x_k)=m. The assertion that “after finitely many steps we either obtain a point x_k ∈ Ω, such that Lu(x_k)<0 or x_k ∉ Ω” is false: a sequence in a bounded domain with increments in a fixed ball can remain inside indefinitely. A correct argument must propagate an open neighborhood of minima, using continuity and the positivity of K_ρ on a neighborhood, or show that {u=m}∩Ω is both open and closed relative to Ω. As written, the strong maximum principle is not established.
[§5.2, Proposition 5.4] The step from ⟨u^-,u^-⟩_ρ = 0 to u^- ≡ 0 is too quick. The hypotheses only give K_ρ ≥ 0 pointwise and (H0). Zero seminorm means u^- is constant on each connected component of the graph whose edges lie in the positive support of K_ρ; it does not automatically force u^-=0 unless every point of Ω is connected to Ω^c through the support of K_ρ. A nondegeneracy statement about K_ρ (for example, positivity on a neighborhood of 0, or a Poincaré-type inequality for the kernel support) is needed. Without it, the weak maximum principle is not proven under the stated assumptions.

minor comments (3)

[Proposition 4.2 and §4.2] There is a sign inconsistency: Proposition 4.2 defines m_{K_ρ}(ξ) as \h\K_ρ(ξ) = -4π²|ξ|²Q̂_ρ² = -|λ_ρ(ξ)|², while Proposition 4.12, Corollary 4.17, and the class condition (3.5) require m_{K_ρ} = |λ_ρ(ξ)|². This notation should be made consistent.
[Theorem 3.4] Typo: “If If lim...” should read “If lim...”. Several other typos appear (“Furtheromre”, “Puting all togheter”, “on of the the authors”); a careful proofreading pass is needed.
[§4.1, proof of Theorem 4.9] The computation of the symbol uses m_{K_ρ}(ξ)=K̂_ρ(0)-K̂_ρ(ξ) and then identifies this with -K̂_ρ(ξ). For kernels with nonintegrable singularity at 0 this is formal; for compactly supported kernels with integrable singularity the term K̂_ρ(0) does not vanish. The proof should be written in terms of the class symbol m_{K_ρ}(ξ)=∫(1-cos(2πξ·h))K_ρ(h)dh to avoid this ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main ρ-Laplacian results follow by direct Fourier computation and do not reduce to their inputs.

full rationale

The paper's central derivation is computational, not self-referential. Δ_ρ is defined as div_ρ(D_ρu) (Definition 4.1), and Proposition 4.2 derives directly by Fourier transform that its symbol is −4π²|ξ|²Q̂_ρ(ξ)² and that the convolution kernel is K_ρ = −D Q_ρ * D Q_ρ. The identification L_ρ = −Δ_ρ in Proposition 4.12 is the same direct computation, and the maximum principle in Proposition 5.2 uses only the sign of K_ρ and the integral representation. The road to Theorem 4.18 goes through Lemma 4.15/4.16, quoting [26, Lemma 2.6] for estimates on Q̂_ρ on |ξ| ≥ 1/ε, and Corollary 4.17 then gives |ξ|^{2s} comparability on that frequency range. There is no fitted parameter renamed as a prediction, no uniqueness claim imported from the authors' prior work, and no hypothesis engineered so that the conclusion is true by construction. The paper does cite works with overlapping authors ([14], [13]), but these supply background nonlocal-gradient framework, not the new conclusion; the load-bearing symbol estimate is cited to [26], whose authors do not overlap with the present paper. The genuine weakness is a correctness gap rather than circularity: Corollary 4.17 only proves the symbol lower bound for |ξ| ≥ 1/ε, while membership in K_s(λ,Λ) requires comparability for all ξ ∈ R^n. For compactly supported ρ, Q_ρ is integrable, so Q̂_ρ(0)>0 and m_Kρ(ξ) ∼ c|ξ|² as ξ→0, which is not comparable to |ξ|^{2s} for s<1. Thus Theorem 4.18 is at least incomplete as stated, but this is a missing low-frequency hypothesis, not a derivation that reduces to its own inputs. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities or free parameters. The load-bearing assumptions are the unverified pointwise positivity (4.6) and the unstated low-frequency symbol comparability needed for membership in K_s; the latter is contradicted by the paper's own truncated-kernel examples.

assumptions (5)

domain assumption (H0): ρ radial, ρ∈L^1_loc, ∫min{1,|x|^{-1}}ρ<∞, inf_{B(0,ε)}ρ>0
Standing hypothesis for the nonlocal gradient framework, taken from [14].
domain assumption (H1)–(H4): near-zero control on ρ (decreasing f_ρ, smoothness, fractional comparability)
Assumed for Poincaré inequalities, embeddings, and for the estimates on Q̂ρ used in Theorem 4.18.
ad hoc to paper (4.6): Kρ(x)≥0 pointwise for x≠0
Introduced without proof or examples; not implied by (H0). Needed for the bilinear form, the Gagliardo spaces, and every maximum principle.
ad hoc to paper Implicit global comparability of m_{Kρ}(ξ) with |ξ|^{2s} for all ξ, including |ξ|≤1/ε
Theorem 4.18 requires the K_s symbol condition at all frequencies. Corollary 4.17 only proves bounds for |ξ|≥1/ε. The low-frequency behavior requires ρ to have a heavy tail; fails for compactly supported kernels such as the truncated Riesz gradient.
ad hoc to paper Dρu=0 in H^{ρ,2}_0(Ω) implies u=0 (a Poincaré-type inequality)
Proposition 5.4's proof concludes u⁻≡0 from ⟨u⁻,u⁻⟩_ρ=0. Under (H0) alone, a zero nonlocal seminorm does not force the function to vanish without a Poincaré inequality, which is not assumed in the proposition.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Nonlocal gradient, the nonlocal Laplacian and maximum principles." pith.science (2026). https://pith.science/paper/LO2C6SHA

@misc{pith2026260713161,
  author       = {Pith},
  title        = {Pith review of: Nonlocal gradient, the nonlocal Laplacian and maximum principles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LO2C6SHA}},
  note         = {Machine review of arXiv:2607.13161}
}

abstract

We study the nonlocal $\rho$-Laplacian, defined as the composition of the nonlocal divergence and gradient operators associated with a general radial kernel $\rho$: $\Delta_\rho u=\mbox{div}_\rho\left(D_\rho u\right)$. Our first main contribution is to establish a precise connection between this operator and the class of integro-differential elliptic operators studied by Fern\'andez-Real and Ros-Oton (\cite{FernandezRos}), identifying explicit conditions on the kernel $\rho$ that guarantee membership in this class. Our second main contribution concerns maximum and comparison principles for the $\rho$-Laplacian. We establish both a strong and a weak maximum principle under conditions on $\rho$ that are strictly weaker than those required for membership in the integro-differential class, thereby covering a genuinely broader family of operators. The results require only minimal assumptions on the kernel, and in particular do not rely on any fractional-type comparability condition.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 1 canonical work pages

[1]

Abantangelo, S

N. Abantangelo, S. Dipierro and E. Valdinoci. A Gentle Invitation to the Fractional World. Springer Cham, 2025

2025
[2]

Acosta and J

G. Acosta and J. P. Borthagaray. A Fractional Laplace Equation: Regularity of Solutions and Finite Element Approximations.SIAM Journal on Numerical Analysis, 55(2):472–495, 2017

2017
[3]

S. Almi, M. Caponi, M. Friedrich and F. Solombrino. A fractional approach to strain- gradient plasticity: beyond core-radius of discrete dislocations.Mathematische An- nalen, 391(3), 4063-4115, 2025

2025
[4]

Almi, et al

S. Almi, et al. Riesz fractional gradient functionals defined on partitions: nonlocal- to-local variational limits. arXiv preprint arXiv:2510.04881 (2025)

arXiv 2025
[5]

J. C. Bellido, J. Cueto, and G. García-Sáez. Compact embeddings of Bessel potential spaces. Preprint, arXiv:2506.01677, 2025

arXiv 2025
[6]

J.C.Bellido, J.CuetoandC.Mora-Corral, FractionalPiolaidentityandpolyconvexity in fractional spaces,Ann. I. H. Poincaré – AN, 37 (2020), pp. 955–981. 22

2020
[7]

Bellido, J

J.C. Bellido, J. Cueto and C. Mora-Corral,Γ-convergence of polyconvex functionals involvings-fractional gradients to their local counterparts,Calc. Var. Partial Differ- ential Equations, 60 (2021)

2021
[8]

J. C. Bellido, J. Cueto, and C. Mora-Corral. Non-local gradients in bounded domains motivated by continuum mechanics: Fundamental theorem of calculus and embed- dings.Advances in Nonlinear Analysis, vol. 12, no. 1, pp. 20220316, 2023

2023

Show all 51 references

[9]

J. C. Bellido, J. Cueto, and C. Mora-Corral. Eringen’s model via linearization of nonlocal hyperelasticity.Mathematics and Mechanics of Solids, 29(4):686–703, 2023

2023
[10]

Bellido, J

J.C. Bellido, J. Cueto, Javier and C. Mora-Corral. Minimizers of nonlocal polycon- vex energies in nonlocal hyperelasticity.Advances in Calculus of Variations, 17(3), 1039:1055, 2024

2024
[11]

J. C. Bellido, J. Cueto, M. D. Foss, et al. Nonlocal Green Theorems and Helmholtz Decompositions for Truncated Fractional Gradients.Applied Mathematics and Opti- mization, 90:16, 2024

2024
[12]

J. C. Bellido and G. García-Sáez. Bessel Potential Spaces and Complex Interpolation: Continuous embeddings. 2025. arXiv:2503.04310 [math.FA].https://arxiv.org/ abs/2503.04310

2025
[13]

J. C. Bellido and G. García-Sáez A general theory of nonlocal elasticity based on nonlocal gradients and connections with Eringen’s model. Preprint, arXiv:2601.19797, 2026

2026
[14]

J. C. Bellido, C. Mora-Corral, and H. Schönberger. Nonlocal gradients: Fundamental theorem of calculus, Poincaré inequalities and embeddings.Journal of the London Mathematical Society, 112(2), e70277, 2025

2025
[15]

Bonito, J

A. Bonito, J. P. Borthagaray, R. H. Nochetto, E. Otárola, and A. J. Salgado. Nu- merical methods for fractional diffusion.Computing and Visualization in Science, 19(5–6):19–46, 2018

2018
[16]

Borthagaray and N

J.P. Borthagaray and N. de León. A mixed formulation for the fractional Poisson problem. Preprint, arXiv:2509.18348, 2025

2025
[17]

J. P. Borthagaray, L. M. Del Pezzo and J. C. R. Niño . Besov regularity of solutions to the Dirichlet problem for the Bessel(p, s)-Laplacian. arXiv preprint arXiv:2603.05298, 2026

2026
[18]

Campos and J.F

P.M. Campos and J.F. Rodrigues. On fractional and classical hyperbolic obstacle-type problems.Discrete Contin. Dyn. Syst. Ser. S, 16(12):3813–3836, 2023

2023
[19]

Campos and J.F

P.M. Campos and J.F. Rodrigues. Unilateral problems for quasilinear operators with fractional Riesz gradients, 2023

2023
[20]

Campos and J.F

P.M. Campos and J.F. Rodrigues. An implicit two-phase obstacle-type problem for thep-Laplacian with the fractional gradient, 2024. 23

2024
[21]

Campos and G

P.M. Campos and G. García-Sáez. Generalized Sobolev-Orlicz spaces based on the Riesz fractional gradient as interpolation and potential spaces. Preprint, arXiv:2606.17770, 2026

2026 arXiv
[22]

Carrero, A

L. Carrero, A. Quaas and A. Zuniga. Existence of solutions to a quasilinear nonlocal PDE. Calc. Var. (64), 224, 2025

2025
[23]

Caffarelli and L

L. Caffarelli and L. Silvestre, An extension problem related to the fractional Laplacian, Comm. Partial Differential Equations, 32: 1245–1260, 2007

2007
[24]

G. E. Comi and G. Stefani, A distributional approach to fractional Sobolev spaces and fractional variation: existence of blow-up,J. Funct. Anal., 277 (2019), pp. 3373–3435

2019
[25]

Cueto, C

J. Cueto, C. Kreisbeck, and H. Schönberger. A variational theory for integral func- tionals involving finite-horizon fractional gradients.Fractional Calculus and Applied Analysis, vol. 26, pp. 2001–2056, 2023

2001
[26]

Cueto, C

J. Cueto, C. Kreisbeck, and H. Schönberger.Γ-convergence involving nonlocal gradi- ents with varying horizon: Recovery of local and fractional models.Nonlinear Analy- sis: Real World Applications, 85:104371, 2025

2025
[27]

Cueto and Joshua M

J. Cueto and Joshua M. Siktar. Localization for nonlocal gradient-based optimal control problems. arXiv preprint arXiv:2605.09220 (2026)

2026 arXiv
[28]

D’Elia and M

M. D’Elia and M. Gunzburger, The fractional Laplacian operator on bounded domains as a special case of the nonlocal diffusion operator,Comput. Math. Appl., 66: 1245– 1260, 2013

2013
[29]

D’Elia, M

M. D’Elia, M. Gulian, H. Olson, and G. E. Karniadakis, Towards a unified theory of fractional and nonlocal vector calculus,Fract. Calc. Appl. Anal., 24:1301–1355, 2021

2021
[30]

Du.Nonlocal Modelling, Analysis, and Computa- tion.SIAM

Q. Du.Nonlocal Modelling, Analysis, and Computa- tion.SIAM. doi = 10.1137/1.9781611975628.fm, URL = https://epubs.siam.org/doi/abs/10.1137/1.9781611975628.fm, 2019

2019 doi
[31]

Q. Du, M. Gunzburger, R.B. Lehoucq, and K. Zhou A nonlocal vector calculus, non- local volume-constrained problems, and nonlocal balance laws,Math. Models Methods Appl. Sci., 23: 493–540, 2013

2013
[32]

D. Engl, A. Molchanova and H. Schönberger. Derivation of variational mem- brane models in the context of anisotropic nonlocal hyperelasticity. arXiv preprint arXiv:2602.17278, 2026

2026
[33]

Fernández-Real and X

X. Fernández-Real and X. Ros-Oton.Integro-Differential Elliptic Equations. Birkhäuser, Cham, 2024

2024
[34]

G. Foghem. Stability of complement value problems for p-Lévy operators.Nonlinear Differ. Equ. Appl, 32(1), 2025

2025
[35]

Gambera, S.A

L. Gambera, S.A. Marano and D. Motrenau. Dirichlet problems with fractional com- peting operators and fractional convection.Fractional Calculus and Applied Analysis, 27:2203–2218, 2024. 24

2024
[36]

García-Sáez

G. García-Sáez. Fractional weighted Sobolev spaces associated to the Riesz fractional gradient. arXiv preprint, arXiv:2512.09575, 2025

2025
[37]

Grafakos.Classical Fourier Analysis

L. Grafakos.Classical Fourier Analysis. 2nd ed., volume 249 of Graduate Texts in Mathematics, Springer, New York, 2008

2008
[38]

Kreisbeck and H

C. Kreisbeck and H. M. J. Schönberger Quasiconvexity in the fractional calculus of variations: Characterization of lower semicontinuity and relaxation.Nonlinear Analysis, 2022

2022
[39]

Kassmann

M. Kassmann. A priori estimates for integro-differential operators with measurable kernels.Calculus of Variations and Partial Differential Equations, 34(1):1–21, 2009

2009
[40]

Kassmann, M

M. Kassmann, M. Rang, and R. W. Schwab. Integro-Differential Equations with Non- linear Directional Dependence.Indiana University Mathematics Journal, 63(5):1467– 1498, 2014

2014
[41]

Kreisbeck and H

C. Kreisbeck and H. M. J. Schönberger. Non-constant functions with zero nonlocal gradient and their role in nonlocal Neumann-type problems.Nonlinear Analysis, 2024

2024
[42]

Kwaśnicki

M. Kwaśnicki. Ten equivalent definitions of the fractional Laplace operator. Fractional Calculus and Applied Analysis, 20(1), 7-51, 2017

2017
[43]

Mengesha and D

T. Mengesha and D. Spector, Localization of nonlocal gradients in various topologies, Calc. Var. Partial Differential Equations, 52:253–279, 2015

2015
[44]

Ros-Oton and J

X. Ros-Oton and J. Serra, The Dirichlet problem for the fractional Laplacian: regu- larity up to the boundary,J. Math. Pures Appl.101(9):275–302, 2014

2014
[45]

Ros-Oton, Nonlocal elliptic equations in bounded domains: a survey,Publ

X. Ros-Oton, Nonlocal elliptic equations in bounded domains: a survey,Publ. Mat., 60:3–26, 2016

2016
[46]

Schikorra, T

A. Schikorra, T. Shieh and D.E. Spector. Regularity for a fractional p-Laplace equa- tion.Communications in Contemporary Mathematics, 20(1), 2018

2018
[47]

Shieh and D.E

T. Shieh and D.E. Spector. On a new class of fractional partial differential equations I.Advances in Calculus of Variations, 8(4):321–336, 2015

2015
[48]

Shieh and D.E

T. Shieh and D.E. Spector. On a new class of fractional partial differential equations II.Advances in Calculus of Variations, 11(3):289–307, 2018

2018
[49]

Šilhavý, Fractional vector analysis based on invariance requirements (critique of coordinate approaches),Contin

M. Šilhavý, Fractional vector analysis based on invariance requirements (critique of coordinate approaches),Contin. Mech. Thermodyn., 32:207–228, 2020

2020
[50]

M. Šilhavý. Fractional strain tensor and fractional elasticity.Journal of Elasticity, 155:425–447, 2024

2024
[51]

Servadei and E

R. Servadei and E. Valdinoci. Weak and viscosity solutions of the fractional Laplace equation.Publicacions Matemàtiques, 58(1):133–154, 2014. 25 Funding J.C.B. is supported byAgencia Estatal de Investigación(Spain) through grant PID2023- 151823NB-I00 andJunta de Comunidades de...

2014

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

Abantangelo, S

N. Abantangelo, S. Dipierro and E. Valdinoci. A Gentle Invitation to the Fractional World. Springer Cham, 2025

2025

[2] [2]

Acosta and J

G. Acosta and J. P. Borthagaray. A Fractional Laplace Equation: Regularity of Solutions and Finite Element Approximations.SIAM Journal on Numerical Analysis, 55(2):472–495, 2017

2017

[3] [3]

S. Almi, M. Caponi, M. Friedrich and F. Solombrino. A fractional approach to strain- gradient plasticity: beyond core-radius of discrete dislocations.Mathematische An- nalen, 391(3), 4063-4115, 2025

2025

[4] [4]

Almi, et al

S. Almi, et al. Riesz fractional gradient functionals defined on partitions: nonlocal- to-local variational limits. arXiv preprint arXiv:2510.04881 (2025)

arXiv 2025

[5] [5]

J. C. Bellido, J. Cueto, and G. García-Sáez. Compact embeddings of Bessel potential spaces. Preprint, arXiv:2506.01677, 2025

arXiv 2025

[6] [6]

J.C.Bellido, J.CuetoandC.Mora-Corral, FractionalPiolaidentityandpolyconvexity in fractional spaces,Ann. I. H. Poincaré – AN, 37 (2020), pp. 955–981. 22

2020

[7] [7]

Bellido, J

J.C. Bellido, J. Cueto and C. Mora-Corral,Γ-convergence of polyconvex functionals involvings-fractional gradients to their local counterparts,Calc. Var. Partial Differ- ential Equations, 60 (2021)

2021

[8] [8]

J. C. Bellido, J. Cueto, and C. Mora-Corral. Non-local gradients in bounded domains motivated by continuum mechanics: Fundamental theorem of calculus and embed- dings.Advances in Nonlinear Analysis, vol. 12, no. 1, pp. 20220316, 2023

2023