[Paper Review] Energy solutions of KPZ are unique

Massimiliano Gubinelli, Nicolas Perkowski|arXiv (Cornell University)|Aug 31, 2015

Stochastic processes and statistical mechanics39 references22 citations

TL;DR

This paper establishes the uniqueness of energy solutions to the Kardar-Parisi-Zhang (KPZ) equation on the real line, resolving a key open problem in stochastic PDEs. By leveraging a pathwise formulation and a novel martingale problem approach, the authors prove that energy solutions differ from the classical Cole-Hopf solution by an intrinsic $ t/12 $ drift, confirming a long-standing conjecture in KPZ universality and completing the well-posedness framework for the equation.

ABSTRACT

The Kardar-Parisi-Zhang (KPZ) equation is conjectured to universally describe the fluctuations of weakly asymmetric interface growth. Here we provide the first intrinsic well-posedness result for the KPZ equation on the real line by showing that its energy solutions (as introduced by Gonçalves and Jara and later refined by Gubinelli and Jara) are unique. Together with various convergence results already present in the literature, this establishes the weak KPZ universality conjecture for a wide class of models. A remarkable consequence is that the energy solution to the KPZ equation is not equal to the Cole-Hopf solution, but it involves an additional drift $t/12$.

Motivation & Objective

• To establish the intrinsic well-posedness of the KPZ equation on $\mathbb{R}$ by proving uniqueness of energy solutions.
• To resolve the long-standing question of whether energy solutions to the KPZ equation are uniquely defined, despite the equation's ill-posedness due to spatial roughness.
• To confirm that the energy solution is not equivalent to the Cole-Hopf solution, but instead includes an additional $ t/12 $ drift term.
• To complete the proof of weak KPZ universality by combining the uniqueness result with prior convergence results from particle systems.
• To provide a rigorous pathwise formulation of the stochastic Burgers and KPZ equations using a martingale problem approach that avoids ill-defined nonlinearities.

Proposed method

• Adopt a pathwise formulation of the martingale problem for the stochastic Burgers equation (SBE), avoiding the need to define the nonlinear drift $ \partial_x u^2 $ pointwise.
• Use the energy solution framework introduced by Gonçalves and Jara, which defines the solution via a generalized martingale problem with a well-defined distributional drift.
• Apply the Cole-Hopf transformation to relate the KPZ solution $ h_t $ to the solution $ Z_t $ of the linear stochastic heat equation (SHE), with the transformation $ h_t = \lambda^{-1} \log Z_t + \lambda^3 t / 12 $.
• Establish convergence of regularized approximations $ h^L_t $ to the energy solution by analyzing the limiting behavior of the renormalized drift and correction terms.
• Prove key lemmas on the convergence of correction terms: $ R^L_t $ vanishes in $ L^2 $, $ Q^L_t $ converges to a zero-variation process, and the renormalization constant $ K^L $ converges to $ \lambda^2 / 12 $.
• Use Fourier analysis and kernel estimates to compute the limit of $ K^L $, showing it converges to $ 1/12 $ via the sum $ \sum_{k=1}^\infty \frac{1}{k^2} = \frac{\pi^2}{6} $, leading to $ \lambda^2 / 12 $ after normalization.

Experimental results

Research questions

• RQ1 Are energy solutions to the KPZ equation on $\mathbb{R}$ uniquely defined, despite the equation's ill-posedness due to spatial roughness?
• RQ2 Does the energy solution to the KPZ equation coincide with the classical Cole-Hopf solution, or is there a discrepancy due to renormalization?
• RQ3 What is the precise value of the additional drift term that distinguishes the energy solution from the Cole-Hopf solution?
• RQ4 Can the martingale problem formulation for the SBE and KPZ be made rigorous without relying on Markov generators or pointwise definitions of nonlinearities?
• RQ5 Does the convergence of particle systems to the KPZ equation imply universality, and is this universality complete once uniqueness is established?

Key findings

• The energy solution to the KPZ equation on $\mathbb{R}$ is uniquely defined, resolving a fundamental well-posedness question in stochastic PDEs.
• The energy solution differs from the Cole-Hopf solution by an intrinsic $ t/12 $ drift, which arises from the renormalization of the nonlinear term in the SBE.
• The renormalization constant $ K^L $ in the regularized approximation converges to $ \lambda^2 / 12 $ as $ L \to \infty $, with the limit computed via Fourier analysis of the smoothing kernel.
• The correction term $ R^L_t $ vanishes in $ L^2 $-norm as $ L \to \infty $, ensuring the convergence of the regularized martingale problem to the true energy solution.
• The process $ Q^L_t $, which accounts for the renormalization of the quadratic variation, converges in probability to a zero-variation process, confirming the stability of the limit.
• Together with prior convergence results, this uniqueness result completes the proof of the weak KPZ universality conjecture for a wide class of interacting particle systems.