Pith. sign in

REVIEW 1 major objections 5 minor 20 references

A Note on a threshold for temporal regularity of stochastic PDEs

T0 review · 1 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read Temporal regularity of a parabolic SPDE is governed by the spatial regularity of its noise coefficient.

desk verdict Sharp and clean negative answer to a 2016 conjecture; the proof has a repairable gap in one approximation step, not a load-bearing flaw. read the letter →

arxiv 2509.07803 v3 pith:XJEMAN3A submitted 2025-09-09 math.PR

Antonio Agresti , Mark Veraar This is my paper · ORCID
classification math.PR MSC 60H1535B6535R60
keywords temporalregularityparabolicSPDEfractionalSobolev-SlobodeckijspacesstochasticconvolutionvariationalSPDEspowersofoperatorscounterexample
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

This paper studies the linear parabolic SPDE du + Au dt = x dβ, u(0) = 0, where A is a positive, invertible, self-adjoint operator on a Hilbert space and x is a fixed element. It proves a sharp equivalence: for every α in [0, 1/2), the solution u lies in the fractional-time Sobolev space L²(Ω; W^{α,2}(0,T;D(A^{1/2}))) if and only if the spatial coefficient x lies in the domain of the fractional power A^α. The two norms are comparable, so temporal regularity is exactly paid for by spatial regularity of x. This shows that a natural extension of the basic stochastic-parabolic estimate fails for every α > 0: time regularity of the diffusion coefficient does not transfer to the solution. It also provides a counterexample to a conjectured time-regularity property for monotone stochastic evolution equations posed in 2016.

What carries the argument

The proof is carried by the stochastic convolution u(t) = ∫₀ᵗ e^{-(t-r)A} x dβ(r). The main new step is a lower bound on the W^{α,2} seminorm of A^{1/2}u: Itô isometry plus self-adjointness give E||A^{1/2}(u(t)-u(s))||² ≥ (1/2)(x - e^{-2(t-s)A}x, x); substituting this into the Slobodeckij seminorm, the leading term is identified via the standard integral representation of fractional powers as a constant times ||A^α x||², while the remainder is a controlled error. A separate spectral-gap lower bound for the L² part, using the exponential decay e^{-tA} ≤ e^{-δt}, supplies the ||x||² term and turns the lower bound into the full norm equivalence. Approximation by x_n = n(n+A)^{-1}x then extends

What would settle it

Take X = ℓ², A e_n = n² e_n, and x_n = n^{-1}. For α ∈ (0, 1/2), Theorem 1.1 predicts u ∈ L²(Ω; W^{α,2}(0,T;D(A^{1/2}))) exactly when α < 1/4, because ||x||²_{D(A^α)} = ∑ n^{4α-2}. Directly evaluating the Itô-isometry expression per mode, or numerically simulating the first many modes, settles the equivalence: a finite W^{α,2} norm at α = 0.3 would falsify the theorem.

Watch

Extended reading notes

Core claim

For the equation du + Au dt = x dβ, u(0) = 0, with A positive, invertible, and self-adjoint, the paper establishes that for all α ∈ [0, 1/2), u ∈ L²(Ω; W^{α,2}(0,T;D(A^{1/2}))) if and only if x ∈ D(A^α), with (E||u||²_{W^{α,2}(0,T;D(A^{1/2}))})^{1/2} comparable to ||x||_{D(A^α)}. In particular, the estimate (P) does not hold for any α > 0. The same mechanism transfers to the variational setting V ⊂ H ⊂ V': for constant diffusion h ∈ H, u ∈ L²(Ω; W^{α,2}(0,T;V)) if and only if h lies in the complex interpolation space [H, V]_{2α}. This refutes the conjectured time-regularity result from 2016, showing that additional spatial regularity—sometimes even boundary conditions—on the diffusion coeffi

Load-bearing premise

The sharp 'only if' direction requires A to have a spectral gap, so e^{-tA} decays exponentially at rate e^{-δt}; if the spectrum of A accumulates at 0, the term that recovers ||x||² from the L² part of the solution can vanish and the claimed norm equivalence may fail.

Editorial extensions

If this is right

  • For additive spatially constant noise, fractional time smoothness of the solution is exactly equivalent to fractional spatial smoothness of the noise coefficient; no positive α is obtained for free.
  • The estimate (P) fails for every α > 0, so the natural Sobolev-Slobodeckij extension of the classical L² estimate for parabolic SPDEs is false in general.
  • In the variational setting, u ∈ L²(Ω; W^{α,2}(0,T;V)) requires h ∈ [H, V]_{2α}; for concrete operators such as the Dirichlet Laplacian this can mean extra Sobolev regularity or even boundary conditions on the diffusion coefficient.
  • Positive results on temporal regularity in the monotone setting must impose additional spatial regularity on the diffusion operator; Theorem 1.3 shows such conditions are necessary, not mere technical assumptions.

Reading between the lines

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

  • A natural extension, not pursued in the paper: for time-dependent g(t), the same Itô-isometry computation should yield a two-sided comparison involving the W^{α,2} norm of A^{1/2-α}g, replacing the constant coefficient x by a full space-time condition.
  • The failure of (P) suggests numerical time-stepping schemes for SPDEs should not assume that time-smooth noise upgrades the solution's temporal regularity; spatial discretisation must carry that burden.
  • In the monotone variational setting, Theorem 1.3 pinpoints exactly which interpolation spaces [H, V]_{2α} matter, implying that the additional assumptions used in earlier positive results are unavoidable rather than convenient.
  • For the Dirichlet Laplacian, the equivalence turns into explicit Sobolev-space conditions that, for α > 1/4, require boundary behaviour of the noise; this predicts that boundary regularity of the diffusion can drive interior temporal regularity.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

1 major / 5 minor

Summary. The paper studies the linear parabolic SPDE (1.1) with additive constant noise g≡x. Theorem 1.1 establishes, for every α∈[0,1/2), that the solution u belongs to L²(Ω;W^{α,2}(0,T;D(A^{1/2}))) if and only if the diffusion coefficient x belongs to D(A^α), with a two-sided norm equivalence. In particular, the natural persistence estimate (P) fails for every α>0. Theorem 1.3 transfers this result to the variational framework of Breit–Hofmanová and gives a counterexample to their conjectured time-regularity property. The proof uses the stochastic convolution representation, Itô's isometry, and Komatsu's formula for fractional powers, plus a density argument.

Significance. If Theorem 1.1 is correct, it provides a sharp, complete characterization of temporal regularity for constant additive noise, and it rigorously refutes a plausible but overly optimistic conjecture in the variational SPDE literature. The main lower-bound computation is clean and fully explicit: the seminorm of A^{1/2}u in W^{α,2} controls ||A^α x|| by a spectral/Komatsu identity, with a remainder that decays as T→∞. The paper also gives concrete examples (shifted Laplacian, Dirichlet Laplacian) and computes the resulting Sobolev characterisation, which is useful. The only substantive gap I found is a density step that is easily repairable; the spectral-gap assumption is harmless because for positive invertible self-adjoint A the constant C_{δ,T} in (2.3) is positive for every T>0 (its derivative is (1−e^{-2δT})/2).

major comments (1)
  1. [§2.1, approximation step after (2.2)] The proof of the converse asserts that u_n→u in L²(Ω;W^{α,2}(0,T;D(A^{1/2−α}))) and then applies (2.2) to differences u_n−u_m. However, (2.2) is an estimate for the D(A^{1/2})-valued Sobolev–Slobodeckij seminorm; convergence in D(A^{1/2−α}) does not control that seminorm, so the limit passage is not justified as written. This is load-bearing because it is exactly how the lower bound is extended from D(A) to D(A^α). The gap is readily fixed: write v=A^{1/2}u and R_n=n(n+A)^{-1}; since R_n is a contraction on X and converges strongly to I, the dominated convergence theorem gives R_n v→v in L²(Ω;W^{α,2}(0,T;X)), and hence u_n→u in L²(Ω;W^{α,2}(0,T;D(A^{1/2}))). With this correction, the rest of the argument goes through.
minor comments (5)
  1. [§2.1, after (2.2)] 'If α=0, then one can use the argument in (2.3) below' — equation (2.3) is introduced later in the proof; a forward pointer or reordering would improve readability.
  2. [§2.1, (2.3)] The positivity of C_{δ,T} is not immediately obvious. It follows because d/dT C_{δ,T} = (1−e^{-2δT})/2 > 0 for all T>0; a one-line comment would help.
  3. [Examples 1.2 and 1.4] The notation W^{2α,2}_0(O) for α∈(1/4,1/2) refers to spaces with zero trace; this is stated earlier but could be repeated for readability.
  4. [Proof of Theorem 1.3] Minor language: 'The operator A naturally restricts on H' should be 'restricts to H'.
  5. [Example 1.2(2)] The display 'u|_{BO}=0' is a typographical artifact for 'u|_{\partial O}=0'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the converse direction is derived from first principles (Itô isometry + Komatsu formula), and the cited forward direction is independent published work.

full rationale

The paper's main theorem (Theorem 1.1) claims an iff between temporal regularity of the solution and spatial fractional regularity of the constant diffusion coefficient. The forward direction ('if x∈D(A^α) then u∈L^2(W^{α,2}(D(A^{1/2})))') is imported from [14], a published theorem of van Neerven–Veraar–Weis. Although co-authored by Veraar, this is a self-citation, but it is independent support: its assumptions (e.g., BIP/self-adjointness) do not include the target result, and it does not contain the converse or the failure of (P). The converse—the main contribution—is derived directly in §2.1: Itô isometry, self-adjointness, the fundamental theorem of calculus, and Komatsu's formula for fractional powers yield the lower bound (2.2). No parameter is fitted, and no quantity is defined in terms of the conclusion. The norm equivalence follows from the same estimates plus the spectral-gap lower bound (2.3). The only notable issue is a proof gap, not a circularity: the approximation step asserts convergence in L^2(Ω;W^{α,2}(D(A^{1/2-α}))) and then applies (2.2), which requires convergence in D(A^{1/2}). This is repairable because R_n=n(n+A)^{-1} is a contraction on D(A^{1/2}) and A^{1/2}u_n=R_n A^{1/2}u, so dominated convergence gives the stronger convergence needed. Thus the derivation is self-contained in its essential converse direction, and the citation to [14] is real evidence rather than load-bearing circularity. Score 0.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper introduces no free parameters and no new entities. It combines standard spectral calculus and stochastic integration with two external results from the literature: the stochastic maximal regularity theorem for the forward direction, and interpolation theorems for the variational corollary. The converse is derived from scratch.

assumptions (6)
  • standard math The spectral theorem for unbounded self-adjoint operators and the associated functional calculus for fractional powers.
    Used throughout to define D(A^{1/2}), D(A^α), and the semigroup e^{-tA}; explicitly invoked in Section 1 and in the proof of Theorem 1.1.
  • standard math Itô isometry for Hilbert-space-valued stochastic integrals.
    Used in the proof of (2.2) and (2.3) to compute the second moments of the stochastic convolution.
  • standard math Komatsu's formula for fractional powers of operators (Martínez-Carracedo and Sanz Alix, Theorem 6.1.6).
    Used to identify ∫_0^∞ (x-e^{-sA}x,x)/s^{1+2α} ds with c_α||A^αx||², the leading term of the lower bound.
  • domain assumption The stochastic maximal L^p-regularity theorem of van Neerven, Veraar and Weis [14], which gives (1.5).
    Provides the sufficiency direction x∈D(A^α) ⇒ u∈L²(Ω;W^{α,2}(0,T;D(A^{1/2}))) and the upper norm estimate.
  • standard math Identification of fractional domains with complex interpolation spaces for BIP operators ([10, Theorem 15.3.9]), plus Seeley's interpolation results [15].
    Used in Examples 1.2 and 1.4 and in the proof of Theorem 1.3 to rewrite D(A^α) as [H,V]_{2α} with explicit Sobolev space descriptions.
  • domain assumption Variational SPDE existence and uniqueness theory ([12, Chapter 4]) and the identification D(A^{1/2})=V ([16, Proposition 1.10]).
    Used in the proof of Theorem 1.3 to identify the variational solution with the stochastic convolution and to transfer Theorem 1.1 to the Gelfand triple setting.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Note on a threshold for temporal regularity of stochastic PDEs." pith.science (2026). https://pith.science/paper/XJEMAN3A

@misc{pith2026250907803,
  author       = {Pith},
  title        = {Pith review of: A Note on a threshold for temporal regularity of stochastic PDEs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XJEMAN3A}},
  note         = {Machine review of arXiv:2509.07803}
}
abstract

We consider solutions to linear parabolic SPDEs of the form \[ \mathrm{d} u(t) + A u(t)\, \mathrm{d} t = g(t)\, \mathrm{d} \beta, \qquad u(0)=0, \] where $A$ is a positive, invertible, and self-adjoint operator on a Hilbert space $X$, $\beta$ is a one-dimensional Brownian motion, and $g(t)\equiv x\in X$. We show that, for all $\alpha\in [0,\frac{1}{2}),$ \[ u\in L^2(\Omega;W^{\alpha,2}(0,T;\mathsf{D}(A^{1/2}))) \quad \text{ if and only if }\quad x\in \mathsf{D}(A^{\alpha}). \] In particular, there is a lack of persistence of temporal regularity from the diffusion coefficient $g$ to the solution, and additional spatial regularity is required to improve time regularity. In particular, this provides a counterexample to a conjectured time-regularity property for monotone stochastic evolution equations posed by D. Breit and M. Hofmanov\'a in [C. R. Math. Acad. Sci. Paris 354 (2016), 33-37].

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 19 canonical work pages

  1. [1]

    Agresti and M.C

    A. Agresti and M.C. Veraar. The critical variational setting for stochastic evolution equations.Probab. Theory Related Fields, 188(3-4):957–1015, 2024

    work page 2024

  2. [2]

    Agresti and M.C

    A. Agresti and M.C. Veraar. Nonlinear SPDEs and maximal regularity: An extended survey.arXiv preprint arXiv:2501.18561, 2025. To appear in Nonlinear Differential Equations and Applications

    arXiv 2025

  3. [3]

    Bergh and J

    J. Bergh and J. L¨ ofstr¨ om.Interpolation spaces. An introduction. Springer-Verlag, Berlin, 1976. Grundlehren der Mathematischen Wissenschaften, No. 223

    work page 1976

  4. [4]

    Breit and M

    D. Breit and M. Hofmanov´ a. On time regularity of stochastic evolution equations with monotone coefficients. C. R. Math. Acad. Sci. Paris, 354(1):33–37, 2016

    work page 2016

  5. [5]

    Breit, M

    D. Breit, M. Hofmanov´ a, and S. Loisel. Space-time approximation of stochasticp-Laplace-type systems.SIAM J. Numer. Anal., 59(4):2218–2236, 2021

    work page 2021

  6. [6]

    Da Prato

    G. Da Prato. Regularity results of a convolution stochastic integral and applications to parabolic stochastic equations in a Hilbert space.Confer. Sem. Mat. Univ. Bari, (182):17, 1982

    work page 1982

  7. [7]

    Da Prato and J

    G. Da Prato and J. Zabczyk.Stochastic equations in infinite dimensions, volume 152 ofEncycl. Math. Appl. Cambridge: Cambridge University Press, 2nd ed. edition, 2014

    work page 2014

  8. [8]

    Hyt¨ onen, J.M.A.M

    T.P. Hyt¨ onen, J.M.A.M. van Neerven, M.C. Veraar, and L. Weis.Analysis in Banach spaces. Vol. I. Martin- gales and Littlewood-Paley theory, volume 63 ofErgebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge. Springer, 2016

    work page 2016

Show all 20 references
  1. [9]

    Hyt¨ onen, J.M.A.M

    T.P. Hyt¨ onen, J.M.A.M. van Neerven, M.C. Veraar, and L. Weis.Analysis in Banach spaces. Vol. II. Prob- abilistic Methods and Operator Theory., volume 67 ofErgebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge.Springer, 2017

    2017

  2. [10]

    Hyt¨ onen, J.M.A.M

    T.P. Hyt¨ onen, J.M.A.M. van Neerven, M.C. Veraar, and L. Weis.Analysis in Banach spaces. Vol. III. Har- monic analysis and spectral theory, volume 76 ofErgebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge. Springer, 2023

    2023

  3. [11]

    Krylov and B.L

    N.V. Krylov and B.L. Rozovski ˘ ı. Stochastic evolution equations. InCurrent problems in mathematics, Vol. 14 (Russian), pages 71–147, 256. Akad. Nauk SSSR, Vsesoyuz. Inst. Nauchn. i Tekhn. Informatsii, Moscow, 1979

    1979

  4. [12]

    Liu and M

    W. Liu and M. R¨ ockner.Stochastic partial differential equations: an introduction. Universitext. Springer, 2015

    2015

  5. [13]

    Mart ´ ınez Carracedo and M

    C. Mart ´ ınez Carracedo and M. Sanz Alix.The theory of fractional powers of operators, volume 187 ofNorth- Holland Math. Stud.Amsterdam: Elsevier, 2001

    2001

  6. [14]

    van Neerven, M.C

    J.M.A.M. van Neerven, M.C. Veraar, and L.W. Weis. Stochastic maximalL p-regularity.Ann. Probab., 40(2):788–812, 2012

    2012

  7. [15]

    R. Seeley. Interpolation inL p with boundary conditions.Studia Math., 44:47–60, 1972

    1972

  8. [16]

    Taylor.Partial differential equations

    M.E. Taylor.Partial differential equations. I: Basic theory, volume 115 ofAppl. Math. Sci.Springer, 3rd corrected and expanded edition, 2023

    2023

  9. [17]

    Triebel.Theory of function spaces, volume 78 ofMonographs in Mathematics

    H. Triebel.Theory of function spaces, volume 78 ofMonographs in Mathematics. Birkh¨ auser Verlag, Basel, 1983

    1983

  10. [18]

    Wichmann.Regularity and Numerics for stochastic p-Laplace and symmetric p-Stokes systems

    J. Wichmann.Regularity and Numerics for stochastic p-Laplace and symmetric p-Stokes systems. PhD thesis, Dissertation, Bielefeld, Universit¨ at Bielefeld, 2022, 2022

    2022

  11. [19]

    Wichmann

    J. Wichmann. On temporal regularity of strong solutions to stochasticp-Laplace systems.SIAM J. Math. Anal., 55(4):3713–3730, 2023

    2023

  12. [20]

    Wichmann

    J. Wichmann. Temporal regularity of symmetric stochasticp-Stokes systems.J. Math. Fluid Mech., 26(2):Paper No. 20, 28, 2024. 8 ANTONIO AGRESTI AND MARK VERAAR Department of Mathematics Guido Castelnuovo, Sapienza University of Rome, P.le Aldo Moro 5, 00185 Rome, Italy Email ad...

    2024

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.