Positive probability of explosion for stochastic heat equation with superlinear accretive reaction term and polynomially growing multiplicative noise

Michael Salins; Yuyang Zhang

arxiv: 2605.11319 · v2 · pith:5MZOFUMDnew · submitted 2026-05-11 · 🧮 math.PR

Positive probability of explosion for stochastic heat equation with superlinear accretive reaction term and polynomially growing multiplicative noise

Michael Salins , Yuyang Zhang This is my paper

Pith reviewed 2026-05-13 01:34 UTC · model grok-4.3

classification 🧮 math.PR

keywords stochastic heat equationfinite time explosionmild solutionsmultiplicative noisesuperlinear reactionspace-time white noiseexplosion probability

0 comments

The pith

Mild solutions to the stochastic heat equation explode with positive probability for beta in (1,3) with gamma in (beta/2, (beta+3)/4) or beta greater than 1 with gamma up to beta/2.

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

The paper proves that mild solutions of the stochastic heat equation with superlinear reaction term of power beta and multiplicative noise coefficient growing like u to the power gamma on the periodic interval can explode in finite time with positive probability. This holds in the regime where beta is between 1 and 3 and gamma lies between beta over 2 and (beta plus 3) over 4, as well as when beta exceeds 1 and gamma is at most beta over 2. A sympathetic reader cares because the result maps out when the nonlinear drift overcomes the noise to produce blow-up, clarifying the boundary between global existence and explosion in parabolic SPDEs. It refines earlier work by covering an intermediate parameter range where the competition between the terms was previously unresolved.

Core claim

The central claim is that for the equation partial u over partial t equals partial squared u over partial x squared plus u to the beta plus sigma(u) dot W, with sigma(u) approximately u to the gamma near infinity, mild solutions explode with positive probability if beta is in (1,3) and gamma in (beta/2, (beta+3)/4), or if beta exceeds 1 and gamma is in (0, beta/2]. The result is obtained on the periodic domain with space-time white noise and provides a partial characterization of explosion behavior in the intermediate regime.

What carries the argument

Mild solutions defined up to the potential explosion time, whose existence allows direct analysis of the probability that the solution reaches infinity in finite time when the reaction and noise exponents satisfy the stated inequalities.

Load-bearing premise

The noise coefficient behaves like a power of the solution near infinity, together with the existence of mild solutions up to the potential explosion time on the periodic interval.

What would settle it

A numerical simulation of a spatial discretization of the equation for parameters inside the claimed range, such as beta equal to 2 and gamma equal to 1.1, showing that the solution remains finite for all time with probability one would falsify the positive probability of explosion.

Figures

Figures reproduced from arXiv: 2605.11319 by Michael Salins, Yuyang Zhang.

**Figure 1.** Figure 1: Explosion regions (A) If β > 1 and γ ∈ (0, β 2 ], we prove that solutions can explode with positive probability in Section 4. (B) If β ∈ (1, 3) and γ ∈ ( β 2 , β+3 4 ), we prove that solutions can explode with positive probability in Section 5. (C) If β ∈ [0, 1] and γ ∈ [0, 3 2 ] then solutions cannot explode by [20, 30] and comparison principle [14, 21]. (D) If β ≥ 0 and γ > 3 2 , then results of Mueller … view at source ↗

**Figure 2.** Figure 2: Explosion regions Section 4 10 [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

**Figure 3.** Figure 3: Explosion regions Section 5 stochastic convolution integrals in (5.1). Let ϕ be an adapted random field, define Z ϕ (t, x) := Z t 0 Z G(t − s, x − y)ϕ(s, y)W(dyds) (5.2) and Y ϕ (t, x) := Z t 0 Z G(t − s, x − y)ϕ(s, y)dyds. (5.3) Theorem 4 (Theorem 1.2 of [30]) For any p > 6 there exists a constant Cp > 0 such that for any T ∈ [0, 1] and any adapted ϕ, E sup t∈[0,T] sup x∈D |Z ϕ (t, x)| p ≤ CpT p 4 − 3 2E … view at source ↗

read the original abstract

This paper studies the finite time explosion of the stochastic heat equation $\frac{\partial u}{\partial t}(t,x)=\frac{\partial^2}{\partial x^2} u(t,x)+(u(t,x))^{\beta}+\sigma(u(t,x))\dot{W}(t,x)$. We consider an interval $D=[-\pi,\pi]$ under periodic boundary condition where $\dot{W}(t,x)$ is a space-time white noise and $\sigma(u)\approx u^{\gamma}$ near $\infty$. Our results refine existing results by identifying behavior in a previously less understood regime, where we show that if $\beta\in(1,3),\gamma\in(\frac{\beta}{2},\frac{\beta+3}{4})$ or $\beta>1,\gamma\in(0,\frac{\beta}{2}]$ then mild solutions can explode with positive probability. This paper provides a partial characterization of the explosion behavior in an intermediate parameter regime, and contribute to the understanding of the interplay between the drift and diffusion terms.

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

The paper establishes positive-probability explosion for the stochastic heat equation in the intermediate regime β∈(1,3) and γ∈(β/2,(β+3)/4), but the local existence of the mild solution up to the blow-up time needs verification when the noise growth exceeds β/2.

read the letter

The central result is that mild solutions explode with positive probability when the drift power β sits between 1 and 3 and the noise power γ lies in (β/2, (β+3)/4), or when γ is at most β/2 for any β>1. This fills a gap left by earlier work that handled the cases where γ is small or very large relative to β. The authors work on the torus with space-time white noise and assume σ(u) behaves like u^γ at infinity, which is a natural setup for these questions. They appear to use the standard mild formulation with the heat semigroup and localize the solution via a stopping time at which the sup-norm blows up. That approach is reasonable and directly addresses the interplay between the superlinear accretive term and the multiplicative noise. The abstract is clear about the parameter ranges, and the argument seems to avoid circularity or fitted parameters. The math is grounded in existing SPDE techniques rather than inventing new entities. The soft spot is the construction of the mild solution on [0,τ). The coefficients are only locally Lipschitz, so one needs a fixed-point or truncation argument that works in a space compatible with the low Hölder regularity coming from space-time white noise. When γ exceeds β/2 the noise term grows faster, and the a priori estimates for the stochastic convolution may not close without additional care. If those estimates fail to hold uniformly up to the first explosion time, the stopping time τ itself could be ill-defined or zero with positive probability, which would undercut the claim. The paper states the result, but the details of how the local existence is proved for the upper end of the γ interval are the part that requires the closest look. This work is for people already working on explosion criteria for SPDEs. A reader who follows the literature on global existence versus blow-up for semilinear stochastic equations will get direct value from the refined parameter ranges. It is not broad enough for general citation outside that niche, but the result is precise enough that a serious editor should send it to referees rather than desk-reject it. Referees will likely focus on the local-existence step and the tightness of the estimates in the intermediate regime.

Referee Report

1 major / 1 minor

Summary. The paper proves that mild solutions to the stochastic heat equation ∂_t u = ∂_{xx} u + u^β + σ(u) Ẇ on the torus [-π,π] with space-time white noise explode in finite time with positive probability, provided β ∈ (1,3) and γ ∈ (β/2, (β+3)/4), or β > 1 and γ ∈ (0, β/2], where σ(u) ∼ u^γ near infinity. The result refines earlier explosion criteria by covering an intermediate regime in which the noise growth competes with the superlinear drift.

Significance. If the local existence and stopping-time arguments hold, the work supplies a more complete picture of the parameter region separating global existence from finite-time blow-up for SPDEs with locally Lipschitz coefficients. It clarifies the balance between the accretive reaction term and the multiplicative noise in the presence of space-time white noise.

major comments (1)

[§3] §3 (local existence and stopping time): the mild solution is constructed on [0,τ) with τ = inf{t : ||u(t)||_∞ = ∞} via truncation and fixed-point in a ball of the mild integral equation. For γ > β/2 the stochastic convolution term must map into a space where u ↦ u^γ remains locally Lipschitz; given the Hölder regularity <1/4 of the space-time white noise, the a priori estimates used to close the contraction (or to obtain a positive lower bound on τ) are not shown to hold uniformly up to the boundary of the claimed regime γ = (β+3)/4. This is load-bearing for the positive-probability explosion statement.

minor comments (1)

[Theorem 1.1] The notation for the explosion time τ and the precise function space in which the mild solution lives should be stated explicitly in the statement of the main theorem (currently only in the abstract).

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and the detailed comment on the local existence construction. We address the concern point by point below and clarify that the estimates hold uniformly inside the open interval for γ.

read point-by-point responses

Referee: [§3] §3 (local existence and stopping time): the mild solution is constructed on [0,τ) with τ = inf{t : ||u(t)||_∞ = ∞} via truncation and fixed-point in a ball of the mild integral equation. For γ > β/2 the stochastic convolution term must map into a space where u ↦ u^γ remains locally Lipschitz; given the Hölder regularity <1/4 of the space-time white noise, the a priori estimates used to close the contraction (or to obtain a positive lower bound on τ) are not shown to hold uniformly up to the boundary of the claimed regime γ = (β+3)/4. This is load-bearing for the positive-probability explosion statement.

Authors: We appreciate the referee highlighting the need for explicit uniformity. In the proof of local existence (Theorem 3.1), we work in the Hölder space C^α([0,T]×𝕋) with α<1/4 chosen small enough depending on γ. The stochastic convolution is estimated via the heat kernel and BDG inequality, producing the key bound (3.12) whose contraction factor contains the term T^θ with θ=(β+3)/4−γ>0. Because the claimed regime is the open interval γ<(β+3)/4, θ remains strictly positive and we may choose T>0 small enough, uniformly in the truncation level N, to close the contraction mapping and guarantee inf τ_N>0 with positive probability. The boundary γ=(β+3)/4 is deliberately excluded precisely because θ=0 would prevent this uniform control. We will add a short remark after the proof of Theorem 3.1 making the dependence on the gap θ explicit. revision: partial

Circularity Check

0 steps flagged

No circularity: standard SPDE local existence and explosion probability proof

full rationale

The derivation proceeds via the standard mild formulation of the stochastic heat equation on the torus, local existence of mild solutions up to a stopping time via truncation or fixed-point arguments in appropriate Hölder spaces, and then separate probabilistic estimates showing that the explosion time is finite with positive probability under the given parameter ranges for β and γ. These steps rely on classical semigroup theory for the heat kernel, Burkholder-Davis-Gundy inequalities for the stochastic convolution, and comparison or test-function arguments for explosion; none reduce by construction to fitted parameters, self-definitions, or load-bearing self-citations. The paper refines prior results in the literature without invoking uniqueness theorems or ansatzes from the authors' own prior work as the sole justification.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard assumptions from stochastic analysis and the specific growth conditions stated for the coefficients.

axioms (2)

domain assumption Existence of mild solutions to the SPDE up to explosion time
The paper discusses mild solutions and their explosion, assuming they exist under the given conditions.
standard math Standard properties of space-time white noise on the periodic domain
Invoked implicitly for the stochastic integral in the mild formulation.

pith-pipeline@v0.9.0 · 5480 in / 1183 out tokens · 35671 ms · 2026-05-13T01:34:50.623194+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

V(In(t)) = V(In(0)) + ∫ [ε u^β / I^{ε+1} - ε(ε+1) σ(u)^2 / (2 I^{ε+2})] dx ds + N(t∧τ_n^∞) with Hölder comparison ∫ u^β ≥ C (∫ σ(u)^2)^{β/(2γ)} when 2γ ≤ β

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.