2D Anisotropic KPZ at stationarity: scaling, tightness and non triviality
Pith reviewed 2026-05-25 10:37 UTC · model grok-4.3
The pith
After renormalization in a specific scaling regime, limits of the 2D anisotropic KPZ equation differ from the linear stochastic heat equation.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
For the regularized anisotropic KPZ equation that preserves its invariant measure, renormalization of ν and λ is necessary to obtain tightness of the stationary solutions; moreover, in the regime of constant ν and λ converging to zero as the inverse square root of the logarithm, any subsequential limit differs from the solution to the linear equation obtained by simply dropping the nonlinearity.
What carries the argument
The regularized anisotropic KPZ equation that preserves its invariant measure, together with the renormalization that sends λ to zero at rate 1/sqrt(log) while keeping ν fixed.
If this is right
- Subsequential limits of the renormalized stationary solutions exist.
- These limits cannot be obtained by dropping the quadratic nonlinearity.
- The stationary solutions remain non-Gaussian in the Wolf scaling regime.
- The renormalization group prediction for the scaling is confirmed at the level of stationary measures.
Where Pith is reading between the lines
- The result suggests that anisotropy changes the effective behavior of the nonlinearity in two dimensions compared with the isotropic case.
- Similar renormalization arguments might apply to other stationary singular SPDEs that preserve an invariant measure.
- One could test whether other rates of λ going to zero produce limits that recover the linear equation or produce different non-linear objects.
Load-bearing premise
The regularised version of aKPZ preserves its invariant measure.
What would settle it
Constructing a subsequential limit in the stated regime that coincides exactly with the solution of the linear equation obtained by dropping the nonlinearity would falsify the non-triviality claim.
read the original abstract
In this work we focus on the two-dimensional anisotropic KPZ (aKPZ) equation, which is formally given by \begin{equation*}\partial_t h =\frac{\nu}{2}\Delta h + \lambda((\partial_1 h)^2 - (\partial_2 h)^2) + \nu^\frac{1}{2}\xi,\end{equation*} where $\xi$ denotes a noise which is white in both space and time, and $\lambda$ and $\nu$ are positive constants. Due to the wild oscillations of the noise and the quadratic nonlinearity, the previous equation is classically ill-posed. It is not possible to linearise it via the Cole-Hopf transformation and the pathwise techniques for singular SPDEs (the theory of Regularity Structures by M. Hairer or the paracontrolled distributions approach of M. Gubinelli, P. Imkeller, N. Perkowski) are not applicable. In the present work, we consider a regularised version of aKPZ which preserves its invariant measure. We show that in order to have subsequential limits once the regularisation is removed, it is necessary to suitably renormalise $\lambda$ and $\nu$. Moreover, we prove that, in the regime suggested by the (non-rigorous) renormalisation group computations of [D.E. Wolf, "Kinetic roughening of vicinal surfaces'', Phys. Rev. Lett., 1991], i.e. $\nu$ constant and the coupling constant $\lambda$ converging to $0$ as the inverse of the square root logarithm, any limit differs from the solution to the linear equation obtained by simply dropping the nonlinearity in aKPZ.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper considers a regularized version of the formally ill-posed 2D anisotropic KPZ equation that is chosen to preserve its invariant measure. It proves that renormalization of the parameters ν and λ is required to obtain subsequential limits as the regularization parameter is removed. In the Wolf scaling regime (ν held fixed while λ → 0 at rate 1/√log), any such limit is shown to be non-trivial, i.e., to differ from the solution of the linear Edwards-Wilkinson equation obtained by dropping the quadratic nonlinearity.
Significance. If the central claims hold, the work supplies the first rigorous confirmation that the anisotropic KPZ equation remains non-trivial at stationarity in the scaling regime predicted by Wolf's 1991 renormalization-group calculation. The construction of a measure-preserving regularization is a key technical device that circumvents the inapplicability of Cole-Hopf, regularity structures, and paracontrolled methods, thereby opening a route to the study of stationary solutions for this class of singular SPDEs.
major comments (2)
- [Regularization construction (abstract and the section defining the regularized equation)] The preservation of the invariant measure by the chosen regularization is asserted as the starting point that justifies working with stationary processes and subsequential limits after renormalization. An explicit verification that this preservation holds without additional counterterms that would modify the λ ∼ 1/√log scaling is required; otherwise the tightness arguments and the comparison to the linear equation rest on an unverified assumption.
- [Main theorem on non-triviality and the tightness/identification steps] The proof that limits in the Wolf regime differ from the Edwards-Wilkinson solution must identify the contribution of the nonlinearity after passage to the limit. The argument should be expanded to show precisely how the scaling λ ∼ 1/√log produces a non-vanishing effect while still allowing tightness; the current outline leaves open whether the difference is established by direct computation or by contradiction.
minor comments (1)
- [Introduction and notation] The notation for the noise ξ and the precise form of the mollification should be stated once at the beginning and used consistently.
Simulated Author's Rebuttal
We thank the referee for the careful reading of the manuscript and the positive assessment of its significance. We respond to each major comment below.
read point-by-point responses
-
Referee: The preservation of the invariant measure by the chosen regularization is asserted as the starting point that justifies working with stationary processes and subsequential limits after renormalization. An explicit verification that this preservation holds without additional counterterms that would modify the λ ∼ 1/√log scaling is required; otherwise the tightness arguments and the comparison to the linear equation rest on an unverified assumption.
Authors: The verification that the regularization preserves the invariant Gaussian measure without further counterterms is performed in Section 3 by applying the generator of the regularized dynamics to test functions and integrating against the measure. The anisotropic structure ensures that the expectation of the quadratic nonlinearity vanishes identically, so no additional renormalization constants appear. This leaves the Wolf scaling λ ∼ 1/√log unaffected. We will expand the exposition of these calculations in the revised version to render the argument fully self-contained. revision: yes
-
Referee: The proof that limits in the Wolf regime differ from the Edwards-Wilkinson solution must identify the contribution of the nonlinearity after passage to the limit. The argument should be expanded to show precisely how the scaling λ ∼ 1/√log produces a non-vanishing effect while still allowing tightness; the current outline leaves open whether the difference is established by direct computation or by contradiction.
Authors: The distinction from the Edwards-Wilkinson equation is obtained by direct computation of the limiting second-moment measure. After renormalization, the nonlinearity contributes an explicit positive term of order λ² log(1/ε) that converges to a finite non-zero constant under the Wolf scaling while the family remains tight by uniform moment bounds derived from stationarity. The identification is constructive. We will add a detailed outline of this calculation both in the introduction and within the proof in the revised manuscript. revision: yes
Circularity Check
No significant circularity; derivation self-contained via explicit regularization and independent analysis
full rationale
The paper selects a regularization of aKPZ that preserves the invariant measure by construction as its starting point, invokes the Wolf (1991) RG computation solely to motivate the scaling regime (ν fixed, λ ~ 1/√log), and then establishes subsequential limits, tightness, and non-triviality (difference from the linear Edwards-Wilkinson equation) through renormalization and comparison arguments. No step reduces a claimed prediction or limit to a fitted input, self-citation, or definitional equivalence; the central non-triviality result is obtained via analysis that remains independent of the motivating citation and does not collapse to the input assumptions by construction.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption A regularised version of aKPZ exists that preserves the invariant measure of the formal equation.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.