Large field problem in coercive singular PDEs

Ilya Chevyrev, Massimiliano Gubinelli · 2025

Read Large field problem in coercive singular PDEs by Ilya Chevyrev, Massimiliano Gubinelli (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-20716.txt.

Introduction

Ilya Chevyrev and Massimiliano Gubinelli prove advance bounds for rough differential equations driven by very irregular noise. Such equations appear when a smooth rule is shaken by a jagged input that has no clean derivative, making standard estimates break down.

Their only key demand on the smooth part is a holding property, meaning solutions stay controlled when the noise is switched off. From a local version of that property they lift to a global bound by scaling, for a broad class of equations in time only or in space and time. The bounds hold in small regions and do not depend on edge conditions, reducing hard cases to the simpler setting of small noise.

Catalog record

CREATORIlya Chevyrev, Massimiliano Gubinelli
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE4 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-20716

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Large field problem in coercive singular PDEs
Ilya Chevyrev, Massimiliano Gubinelli

Abstract
We derive a priori estimates for singular differential equations of the form ℒ ​ φ = P ⁡ ( φ , ∇ φ ) + f ⁡ ( φ , ∇ φ ) ​ ξ \mathcal{L}\varphi=P(\varphi,\nabla\varphi)+f(\varphi,\nabla\varphi)\xi where P P is a polynomial, f f is a sufficiently well-behaved function, and ξ \xi is an irregular distribution such that the equation is subcritical. The differential operator ℒ \mathcal{L} is either a derivative in time, in which case we interpret the equation using rough path theory, or a heat operator, in which case we interpret the equation using regularity structures. Our only assumption on P P is that solutions with ξ = 0 \xi=0 exhibit coercivity. Our estimates are local in space and time, and independent of boundary conditions. One of our main results is an abstract estimate that allows one to pass from a local coercivity property to a global one using scaling, for a large class of equations. This allows us to reduce the problem of deriving a priori estimates to the case when ξ \xi is small.



1
Introduction

Global aspects of solution theory to stochastic partial differential equations (SPDEs) remain an open problem. In particular, the large field regime—where solutions may become arbitrarily large—poses significant challenges for a priori estimates and existence.

We develop a strategy for a priori estimates based on the fact that the obstruction to such estimates is given by the presence of solutions which blow up at specific space-time points. In the neighbourhood of such points, the solution is large and the equation can be approximated by its scale-invariant version. In the subcritical regime, this corresponds to a semi-classical situation where the noise become small and the equation is dominated by its deterministic part.
If the deterministic equation has a coercive non-linearity then blow-up cannot occur. This intuition can be made rigorous and we show that it leads to a priori estimates for a large class of subcritical singular SPDEs.

We note an intriguing convergence of ideas, where one begins with an example that violates a desired estimate (or property) and, through a scaling argument, arrives at a contradiction by reducing the problem to a much simpler case—typically in a ‘semi-classical’ regime where the noise is negligible.

In particular, we are aware of two other situations where this happens:

•

Symmetry:
In
[
CCHS_2D
,
CCHS_3D
,
BGHZ22
]
the authors use this line of reasoning to prove structural properties of renormalisation, in particular the equivariance under symmetries of the renormalized equation. It were these works that inspired us to tackle the large field problem via a similar strategy.

•

Regularity:
This line of reasoning is used to prove Schauder estimates in a kernel-free way, see e.g.
[
Simon_97
,
Sauer_Smith_25_Schauder
,
Esquivel_Weber_24_fractional
]
.
The idea is to zoom into a space-time point where the desired estimate fails and then use scaling to reduce the problem to a situation where the contributions by the noise are small. In this regime, the desired estimate can be proved by a Liouville property of the linear part of the equation.

In all these cases, we see that a particular property of the SPDE is related to some limit where the noise disappears, either in small scales or for large values of the field, i.e. near an eventual blow-up, or in a regime where a symmetry is broken at the semi-classical level.

It would be interesting to explore this heuristic further especially in connection with geometric aspects of singular SPDEs.

Setting and main results.

We consider a partial differential equation (PDE) of the form

Read the complete text