Temporal properties of the stochastic fractional heat equation with rough dependence in space

Beibei Zhang; Bin Qian

Temporal properties of the stochastic fractional heat equation with rough dependence in space

T0 review · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read The paper proves that the temporal process of the nonlinear stochastic fractional heat equation driven by rough spatial noise behaves, at small scales, exactly like a fractional Brownian motion with index (α+2H−2)/(2α), yielding explicit Kh

desk verdict The temporal approximation theorem is a real extension, but the LIL constants in Corollary 3.1 rest on a decomposition that contradicts the linear solution's covariance when α≠2. read the letter →

arxiv 2607.28167 v2 pith:RT57NTTP submitted 2026-07-30 math.PR

Beibei Zhang , Bin Qian This is my paper · ORCID

classification math.PR MSC 60H1560G1760G22

keywords stochasticfractionalheatequationBrownianmotionlawoftheiteratedlogarithmKhinchin'sLILChung'sroughspatialnoiseLaplaciantemporalincrements

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 establishes that, for the nonlinear stochastic fractional heat equation with space-rough Gaussian noise, the temporal increment at a fixed point equals the coefficient $\sigma(u(t,x))$ times the linear solution's increment, plus a remainder of order $\varepsilon^{(\alpha+2H-2)/(2\alpha)+\eta}$. From this approximation it derives sharp Khinchin and Chung laws of the iterated logarithm for the temporal process $t \mapsto u(t,x)$, with explicit constants. The result matters because it shows the local temporal oscillation of the solution is governed by a single scaling exponent, and that the nonlinearity enters only through the value of $\sigma$ at the point $(t,x)$. If correct, it provides a complete first-order description of the temporal sample paths in the regime of rough spatial dependence and fractional diffusion.

What carries the argument

The load-bearing object is the Gaussian decomposition (1.5): for fixed $x$, the linear solution's temporal process can be written as $\kappa B^{\tilde H/2}_t$ minus an infinitely differentiable remainder. This transfers the Khinchin and Chung LIL constants from a fractional Brownian motion to the linear solution and then, via the approximation theorem, to the nonlinear solution. The approximation theorem itself is proved by decomposing the temporal increment into J0 (initial condition), J1 (historical stochastic convolution), and J2 (short-time stochastic convolution), using a time splitting $s \in [0, t-\varepsilon^\theta]$ and $[t-\varepsilon^\theta, t]$, fractional heat-kernel estimates, and a BDG-type inequality for the rough spati

What would settle it

Compute the exact leading-order variance of the temporal increment of the linear solution $v(t,x)$ in (1.4) for small $\varepsilon$: if $\operatorname{Var}(v(t+\varepsilon,x)-v(t,x))$ does not equal $\kappa^2 \varepsilon^{2\tilde H/2}$ plus smaller-order terms, with $\kappa$ as in (1.6), then the decomposition (1.5) fails. Alternatively, simulate the linear equation at a fixed $x$ and estimate the Khinchin limsup; any deviation from $\kappa$ would falsify the paper's core mechanism.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.1: for every $p \ge 1$ and $\eta$ in the stated interval, the $L^p$ norm of the approximation error — the difference between the temporal increment of the nonlinear solution and the deterministic initial-value term plus $\sigma(u(t,x))$ times the linear temporal increment — is bounded by $c_{3,1} \varepsilon^{(\alpha+2H-2)/(2\alpha)+\eta}$. Using this and the asserted Gaussian decomposition of the linear solution, $t \mapsto \kappa B^{\tilde H/2}_t - v(t,x)$ with infinitely differentiable remainder, the paper derives Corollary 3.1: with probability one, the Khinchin limsup equals $\kappa |\sigma(u(t,x))|$ and the Chung liminf equals $\kappa \lambda_{\tilde H/2}/\tilde H |\sigma(u(t,x))|$, where $\tilde H = (\alpha+2H-2)/\alpha$. Thus the local temporal process

Load-bearing premise

The Gaussian decomposition (1.5) — that the linear temporal process is a scaled fractional Brownian motion plus an infinitely differentiable remainder — is asserted without proof or citation, and it is what transfers the fBm LIL constants to the nonlinear solution; Lemma 5.2's a.s. sup bound is also stated only by 'adapting' a cited work.

Editorial extensions

If this is right

If Theorem 3.1 and Corollary 3.1 are correct, the temporal process at each fixed point has an exact first-order asymptotic: its normalized oscillations converge to those of a fractional Brownian motion with index (α+2H−2)/(2α), and the LIL constants are known explicitly in terms of κ, the small-ball constant λ, and σ(u(t,x)).
The same scaling exponent \tilde H/2 controls both the limsup and liminf rates, so the paper yields a consistent self-similarity index for the temporal process.
At t=0, the Khinchin and Chung constants change by the factor 2^{(1−\tilde H)/2} (via \tilde κ), provided the initial condition is sufficiently Hölder continuous, extending the results to the origin.
The a.s. sup bound in Lemma 5.2, if justified, upgrades the L^p approximation error to a uniform-in-ε almost sure statement, which is what allows the LIL to be stated with probability one.

Reading between the lines

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

The paper treats the Gaussian decomposition (1.5) as an input without proof; if this decomposition were only Hölder rather than C∞, or if κ were miscomputed, the LIL constants would change — so verifying this decomposition directly for the linear equation would be a decisive test of the whole argument.
Because the constant κ depends only on \tilde H and not on α and H separately, the local temporal law may be universal across the allowed parameter region; this could be tested numerically by simulating the linear equation for different (α,H) pairs and checking the same κ.
The dependence on σ(u(t,x)) suggests that at points where σ(u(t,x)) vanishes, the stated LIL degenerates and a different normalization might be needed; the paper does not explore this regime.
The same approximation method could plausibly yield a functional LIL or a Chung-type law for the entire temporal path, since the smooth remainder in (1.5) is negligible at the LIL scale.

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.

Desk Editor's Note

read the letter

Short version: Theorem 3.1 is worth a referee. Corollary 3.1, as stated, is not.

The genuinely new piece is the sharp L^p approximation of temporal increments for the SFHE with α∈(1,2) and H∈((3−α)/4,1/2). That regime was not covered by [32] or [27], and the proof has real content—Proposition 2.2 gives a new kernel estimate, and the exponent bookkeeping in Section 4 mostly holds up. If the paper only claimed (3.1)–(3.2), I would be fairly comfortable.

The problem is the transfer from Theorem 3.1 to Corollary 3.1. Equation (1.5) is asserted, not proved, and the stress-test is right: from the spectral covariance used in the paper,

E[(v(t+ε,x)−v(t,x))²] ~ [2 c_{1,1} Γ((2−2H)/α)/(α rH)] ε^{rH}.

For α=2 this coefficient reduces to Γ(2rH)/Γ(rH) by the usual gamma identity, which is presumably why the formula was imported from [27]. For α≠2 it does not. At α=1.3, H=0.45 the coefficient is about 1.68, while Γ(2rH)/Γ(rH) is about 0.48. So Eq (1.5) cannot hold with the stated κ, and the Khinchin/Chung constants in (3.3) and (3.4) are off by a factor around √3.5. This is not a fillable gap; it is a wrong constant in the main advertised result.

The rest of the soft spots are proportionate. Lemma 5.2 and Lemma 5.3 are borrowed from [27] by “adapting,” and Corollary 3.1’s proof is one sentence. That would be acceptable if the borrowed result had the right constant, but here it doesn’t. The self-citation itself is not a problem.

Bottom line: Theorem 3.1 deserves a serious referee; Corollary 3.1 needs the constant corrected and the decomposition (1.5) either proved or replaced. I would not cite the LIL results as they stand, but I would send the paper to review and require the authors to fix the linear analysis before publication.

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

The proof rests on imported well-posedness ([21]), stochastic-integration framework ([13]), heat-kernel bounds ([1,2,5]), the asserted fBm decomposition (1.5), and standard fBm LIL results. The only novel unproved premise is Lemma 5.2.

assumptions (6)

domain assumption Well-posedness of (1.1) as stated in Theorem 2.1 from Liu–Mao [21].
The paper builds on this external theorem; Condition 2.1 is assumed throughout.
domain assumption Stochastic integration framework for rough spatial noise (Proposition 2.1 from Hu et al. [13]) including isometry and BDG inequality (2.15).
Used throughout the proof; imported from [13].
domain assumption The Gaussian decomposition of the linear solution: t ↦ κB^{\tilde H/2}_t − v(t,x) is smooth on (0,∞) (Eq. (1.5)).
Stated without proof or citation; underpins Lemma 5.3 and Corollary 3.1.
standard math Heat kernel estimates (2.7)–(2.10), (2.11), and Proposition 2.2.
Derived from stable-process kernel bounds in [1,2,5]; used in the main estimates.
standard math Khinchin's and Chung's LIL for fractional Brownian motion and the small-ball constant λ (Li–Shao [20], Monrad–Rootzén [22], Talagrand [29]).
Used in Lemma 5.3 and Corollary 3.1.
ad hoc to paper Lemma 5.2 (maximal a.s. sup estimate for the approximation error) as asserted.
The lemma is stated in the appendix but its proof is only a reference to [27]; this is a load-bearing unproved premise for Corollary 3.1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Temporal properties of the stochastic fractional heat equation with rough dependence in space." pith.science (2026). https://pith.science/paper/RT57NTTP

@misc{pith2026260728167,
  author       = {Pith},
  title        = {Pith review of: Temporal properties of the stochastic fractional heat equation with rough dependence in space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RT57NTTP}},
  note         = {Machine review of arXiv:2607.28167}
}

abstract

This paper investigates the nonlinear stochastic fractional heat equation driven by a Gaussian noise that is white in time and fractional in space with a Hurst parameter $H \in \big(\frac{3-\alpha}{4}, \frac{1}{2}\big)$. Specifically, the driving operator is the fractional Laplacian of order $\alpha/2 \in (1/2, 1)$. We characterize the asymptotic behavior of the temporal increment $u(t+\varepsilon,x)-u(t,x)$ for fixed $t\ge 0$ and $x\in\mathbb{R}$ as $\varepsilon\downarrow 0$. Utilizing these precise asymptotic estimates, we establish Khinchin's and Chung's laws of the iterated logarithm for the temporal process $t \mapsto u(t,x)$.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 1 linked inside Pith

[32]

Wang and Y

R. Wang and Y. Xiao, Temporal properties of the stochastic fractional heat equation with spatially-colored noise. Theor. Probability and Math. Statist. , 110, 121-142 (2024)

2024
[27]

B. Qian, M. Wang, R. Wang and Y. Xiao, Temporal regularity for the nonlinear stochastic heat equation with spatially rough noise. J. Differential Equations , 461, 114097 (2026)

2026
[1]

Assaad, D

O. Assaad, D. Nualart, C. A. Tudor and L. Viitasaari, Quantitative normal approximations for the stochastic fractional heat equation. Stoch. Partial Differ. Equ. Anal. Comput. , 10(1), 223-254 (2022)

2022
[2]

Blumenthal and R

R. Blumenthal and R. Getoor, Some theorems on stable processes. Trans. Amer. Math. Soc. 95, 263-273 (1960)

1960
[3]

Balan, M

R. Balan, M. Jolis and L. Quer-Sardanyons, SPDEs with affine multiplicative fractional noise in space with index 1 4 ă Hă 1 2 . Electron. J. Probab., 20(54), 1-36 (2015)

2015
[4]

Balan, M

R. Balan, M. Jolis and L. Quer-Sardanyons, SPDEs with rough noise in space: H¨ older continuity of the solution. Statist. Probab. Lett., 119, 310-316 (2016)

2016
[5]

Chen and X

Z.-Q. Chen and X. Zhang, Heat kernels and analyticity of non-symmetric jump diffusion semi- groups. Probab. Theory Related Fields, 165(1-2), 267-312 (2016)

2016
[6]

Da Prato and J

G. Da Prato and J. Zabczyk, Stochastic equations in infinite dimensions . Second edition. Cam- bridge University Press, Cambridge (2014)

2014

Show all 33 references

[7]

R. C. Dalang, Extending the martingale measure stochastic integral with applications to spatially homogeneous s.p.d.e.’s. Electron. J. Probab., 4(6), 1-29 (1999)

1999
[8]

R. C. Dalang and L. Quer-Sardanyons, Stochastic integrals for spde’s: a comparison. Expo. Math., 29(1), 67-109 (2011)

2011
[9]

Das, Temporal increments of the KPZ equation with general initial data

S. Das, Temporal increments of the KPZ equation with general initial data. Electron. J. Probab., 29(190), 1-28 (2024)

2024
[10]

Foondun, D

M. Foondun, D. Khoshnevisan and P. Mahboubi, Analysis of the gradient of the solution to a sto- chastic heat equation via fractional Brownian motion. Stoch. Partial Differ. Equ. Anal. Comput. , 3, 133-158 (2015)

2015
[11]

Herrell, R

R. Herrell, R. Song, D. Wu and Y. Xiao, Sharp space-time regularity of the solution to stochastic heat equation driven by fractional-colored noise. Stoch. Anal. Appl. 38(4), 747-768, 2020

2020
[12]

Hu, Some recent progress on stochastic heat equations

Y. Hu, Some recent progress on stochastic heat equations. Acta Math. Sci. Ser. B (Engl. Ed.) , 39(3), 874-914 (2019)

2019
[13]

Y. Hu, J. Huang, K. Lˆ e, D. Nualart and S. Tindel, Stochastic heat equation with rough dependence in space. Ann. Probab., 45(6B), 4561-4616 (2017)

2017
[14]

Y. Hu, J. Huang, K. Lˆ e, D. Nualart and S. Tindel, Parabolic Anderson model with rough de- pendence in space. Computation and combinatorics in dynamics, stochastics and control , 477-498. Abel Symp., 13, Springer, Cham (2018)

2018
[15]

Hu and X

Y. Hu and X. Wang, Stochastic heat equation with general rough noise. Ann. Inst. Henri Poincar´ e Probab. Stat., 58(1) 379-423 (2022)

2022
[16]

Z. M. Khalil and C. A. Tudor, On the distribution and q-variation of the solution to the heat equation with fractional Laplacian. Probab. Math. Statist. 39(2), 315-335 (2019)

2019
[17]

Khoshnevisan, Analysis of stochastic partial differential equations

D. Khoshnevisan, Analysis of stochastic partial differential equations . American Mathematical Soc. (2014)

2014
[18]

Khoshnevisan and M

D. Khoshnevisan and M. Sanz-Sol´ e, Optimal regularity of SPDEs with additive noise. Electron. J. Probab. 28(142), 1-31 (2023)

2023
[19]

Khoshnevisan, J

D. Khoshnevisan, J. Swanson, Y. Xiao and L. Zhang, Weak existence of a solution to a differential equation driven by a very rough fBm. Preprint, arXiv:1309.3613 (2013)

2013 arXiv
[20]

W. V. Li and Q.-M. Shao, Gaussian processes: inequalities, small ball probabilities and applica- tions. In Stochastic Processes: Theory and Methods. Handbook of Statistics , 19, (C.R. Rao and D. Shanbhag, editors), pp. 533–597, North-Holland (2001)

2001
[21]

Liu and L

J. Liu and L. Mao, Nonlinear fractional stochastic heat equation driven by Gaussian noise rough in space. Bull. Sci. Math. , 181, 103207 (2022)

2022
[22]

Monrad and H

D. Monrad and H. Rootz´ en, Small values of Gaussian processes and functional laws of the iterated logarithm. Probab. Theory Related Fields, 101(2), 173-192 (1995)

1995
[23]

Mueller and R

C. Mueller and R. Tribe, Hitting properties of a random string. Electron. J. Probab., 7(10), 1-29 (2002) 13

2002
[24]

Peszat and J

S. Peszat and J. Zabczyk, Stochastic evolution equations with a spatially homogeneous Wiener process. Stochastic Process. Appl., 72(2), 187-204 (1997)

1997
[25]

Peszat and J

S. Peszat and J. Zabczyk, Nonlinear stochastic wave and heat equations. Probab. Theory Related Fields, 116(3), 421-443 (2000)

2000
[26]

Pipiras and M

V. Pipiras and M. S. Taqqu, Integration questions related to fractional Brownian motion. Probab. Theory Related Fields, 118(2), 251-291 (2000)

2000
[28]

Song, SPDEs with colored Gaussian noise: a survey

J. Song, SPDEs with colored Gaussian noise: a survey. Commun. Math. Stat., 6(4), 481-492 (2018)

2018
[29]

Talagrand, Sharper bounds for Gaussian and empirical processes

M. Talagrand, Sharper bounds for Gaussian and empirical processes. Ann. Probab., 22(1), 28-76 (1994)

1994
[30]

C. A. Tudor and Y. Xiao, Sample paths of the solution to the fractional-colored stochastic heat equation. Stoch. Dyn., 17(1), 1750004 (2017)

2017
[31]

Wang, Analysis of the gradient for the stochastic fractional heat equation with spatially-colored noise in R

R. Wang, Analysis of the gradient for the stochastic fractional heat equation with spatially-colored noise in R. Discrete Contin. Dyn. Syst. (B) , 29(6), 2769-2785 (2024)

2024
[33]

Wang and S

R. Wang and S. Zhang, Decompositions of stochastic convolution driven by a white-fractional Gaussian noise. Front. Math. China 16(4), 1063-1073 (2021) Beibei Zhang, School of Mathematics and Statistics, Suzhou University of Technology, Changshu, Jiangsu, 215500, China. Email a...

2021

Pith tools

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

[1] [32]

Wang and Y

R. Wang and Y. Xiao, Temporal properties of the stochastic fractional heat equation with spatially-colored noise. Theor. Probability and Math. Statist. , 110, 121-142 (2024)

2024

[2] [27]

B. Qian, M. Wang, R. Wang and Y. Xiao, Temporal regularity for the nonlinear stochastic heat equation with spatially rough noise. J. Differential Equations , 461, 114097 (2026)

2026

[3] [1]

Assaad, D

O. Assaad, D. Nualart, C. A. Tudor and L. Viitasaari, Quantitative normal approximations for the stochastic fractional heat equation. Stoch. Partial Differ. Equ. Anal. Comput. , 10(1), 223-254 (2022)

2022

[4] [2]

Blumenthal and R

R. Blumenthal and R. Getoor, Some theorems on stable processes. Trans. Amer. Math. Soc. 95, 263-273 (1960)

1960

[5] [3]

Balan, M

R. Balan, M. Jolis and L. Quer-Sardanyons, SPDEs with affine multiplicative fractional noise in space with index 1 4 ă Hă 1 2 . Electron. J. Probab., 20(54), 1-36 (2015)

2015

[6] [4]

Balan, M

R. Balan, M. Jolis and L. Quer-Sardanyons, SPDEs with rough noise in space: H¨ older continuity of the solution. Statist. Probab. Lett., 119, 310-316 (2016)

2016

[7] [5]

Chen and X

Z.-Q. Chen and X. Zhang, Heat kernels and analyticity of non-symmetric jump diffusion semi- groups. Probab. Theory Related Fields, 165(1-2), 267-312 (2016)

2016

[8] [6]

Da Prato and J

G. Da Prato and J. Zabczyk, Stochastic equations in infinite dimensions . Second edition. Cam- bridge University Press, Cambridge (2014)

2014