Boston University
Global existence for stochastic reaction-diffusion equations with constraining potentials
Abstract
dc:description.abstractThis thesis explores a pathwise framework for analyzing a large class of singular stochastic reaction-diffusion equations (SRDEs) of the form \[ \frac{\partial u}{\partial t}(t,x) = \mathcal{A} u(t,x) + f(u(t,x)) + \sigma(u(t,x)) \dot{W}(t,x), \] posed on a bounded, open spatial domain D \subset \mathbb{R}d with initial data ||u0||L\infty(D) < 1. We consider the case where $\mathcal{A}$ is a second-order self-adjoint elliptic operator with Dirichlet boundary conditions, $f$ is a dissipative drift, $\dot{W}$ is a Gaussian noise, and f, σ possess singularities of size β and γ, respectively, near $u = \pm 1$ (e.g., |σ(u)| \le C ( 1 - |u|)-γ near $u = \pm 1$ for some γ \ge 0). \par In SPDEs, classical solutions do not exist because paths are not differentiable and we instead look for so-called mild solutions, which can be understood as integral solutions that are weak PDE solutions via integration against smooth test functions. Assuming a Lipschitz continuous drift and noise coefficient, these mild solutions exist and are unique for all time. However, long-time existence is less straightforward in the presence of highly nonlinear f, σ. We demonstrate that, among other assumptions, γ + 1 < \frac{1-\eta}{2} (β+1) is sufficient to ensure solutions remain within $(-1,1)$ for all time with probability one, which we call global existence, where $\frac{1-\eta}{2}$ is the temporal H\"older regularity exponent of $u(t,x)$. \par The main advantage of this framework lies in bypassing It\^o's formula to interpret SPDE solutions pathwise as solutions to a random PDE. The dynamics of this random PDE control the size of solutions through dissipativity, which are typically unavailable in the stochastic setting. While the framework yields global existence criteria, we also investigate its limitations in the study of finite-time blowup, highlighting the structural obstructions in such global random PDE techniques. Thus, the random PDE framework bridges traditional deterministic control and stochastic analysis while also identifying regimes where such comparisons break down.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Mathematics & Statistics
- Grantor
- Boston University
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ivanhoe, John Nuanjiang
- Advisor dc:contributor.advisor
-
- Salins, Michael
Subjects
dc:subject × 7Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/2144/53739