Spatial covariance of KPZ from flat initial profile

Le Chen, Juan J. Jiménez · 2026

Read Spatial covariance of KPZ from flat initial profile by Le Chen, Juan J. Jiménez (2026) 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-2603-14174.txt.

Introduction

Chen and Jiménez compute the fixed-time spatial covariance of the KPZ equation with a flat starting profile. With Malliavin calculus and a Clark-Ocone form, they find the far-field decay as distance grows. The covariance falls off like a Gaussian in space, with a rate set by time and a factor from the flat heat equation. That shape contrasts with the narrow-wedge case, where decay is a slow power law. The result is the first exact covariance formula of this kind for flat data.

Catalog record

CREATORLe Chen, Juan J. Jiménez
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE73239
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.14174

Browse the full library · About the Tower of Babel

Full text

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

Spatial covariance of KPZ from flat initial profile Le Chen, Juan J. Jiménez Abstract We study the fixed-time spatial covariance of the KPZ equation with flat initial profile. Using Malliavin calculus and a Clark–Ocone representation, we show that as | x | → ∞ |x|\to\infty , Cov ⁡ [ h ⁡ ( t , x ) , h ⁡ ( t , 0 ) ] \Cov[h(t,x),h(t,0)] is governed by a boundary-layer regime near the initial time and satisfies Cov ⁡ [ h ⁡ ( t , x ) , h ⁡ ( t , 0 ) ] ∼ κ ⁡ ( t ) ​ ∫ 0 t p 2 ​ r ​ ( x ) ​ 𝑑 r = 2 ​ κ ​ ( t ) π ​ t 3 / 2 ​ | x | − 2 ​ exp ⁡ ( − x 2 4 ​ t ) , | x | → ∞ , \Cov[h(t,x),h(t,0)]\sim\kappa(t)\int_{0}^{t}p_{2r}(x)\,\mathrm{d}r=\frac{2\kappa(t)}{\sqrt{\pi}}\,t^{3/2}\,|x|^{-2}\exp\!\left(-\frac{x^{2}}{4t}\right),\quad|x|\to\infty, where κ ⁡ ( t ) = ( 𝔼 ⁡ [ Z ​ ( t , 0 ) − 1 ] ) 2 \kappa(t)=\big(\mathbb{E}[Z(t,0)^{-1}]\big)^{2} , Z Z is the flat stochastic heat equation solution, and p t p_{t} is the one-dimensional heat kernel. In sharp contrast with the narrow-wedge regime, where Gu–Pu [ 10 ] *Theorem 1.1 proved that for each fixed t > 0 t>0 , Cov ⁡ [ h nw ​ ( t , x ) , h nw ​ ( t , 0 ) ] ∼ t | x | , | x | → ∞ , \Cov\!\left[h^{\mathrm{nw}}(t,x),h^{\mathrm{nw}}(t,0)\right]\sim\frac{t}{|x|},\qquad|x|\to\infty, the flat initial profile exhibits Gaussian decay, yielding, to the best of our knowledge, the first exact spatial covariance asymptotic for the KPZ equation under flat initial data. We also establish an explicit closed-form formula for the second moment of the continuum directed random polymer partition function. Keywords: KPZ equation, stochastic heat equation, spatial covariance, boundary-layer analysis, Malliavin calculus, Clark–Ocone formula, continuum directed random polymer, normalized Green’s function, inverse moments. Mathematics Subject Classification (2020): 60H15, 60H07, 35R60. 1 Introduction In this paper, we study the Kardar–Parisi–Zhang (KPZ) equation [ 12 ] ( 1.1 ): \partial_{t}h(t,x)=\tfrac{1}{2}\partial_{x}^{2}h(t,x)+\tfrac{1}{2}\big(\partial_{x}h(t,x)\big)^{2}+\xi(t,x),\qquad t>0,\ x\in\mathbb{R}, (1.1) subject to zero initial data , h(0,x)=0 . Its Cole–Hopf solution is defined by h(t,x)=\log Z(t,x) , where Z solves the stochastic heat equation (SHE), \partial_{t}Z(t,x)=\tfrac{1}{2}\partial_{x}^{2}Z(t,x)+Z(t,x)\,\xi(t,x),\qquad t>0,\ x\in\mathbb{R}, (1.2) subject to the flat initial condition : Z(0,x)=u_{0}(x)\coloneqq 1. (1.3) In both ( 1.1 ) and ( 1.2 ), \xi is a space–time white noise on \mathbb{R}_{+}\times\mathbb{R} on some probability space (\Omega,\mathcal{F},\mathbb{P}) , and (\mathcal{F}_{t})_{t\geq 0} denotes the complete filtration generated by \xi . The KPZ equation (and related polymer models) has been studied extensively; we refer to [ 9 ] for a broad overview. Our focus is the fixed-time spatial covariance of the height field (under flat initial data, the law is stationary in x ). In the narrow-wedge setting, Gu–Pu [ 10 ] determine the large- |x| asymptotic of \Cov[h(t,x),h(t,0)] and show that the dominant contribution comes from a boundary-layer geometry near the initial time. Under flat initial data, integrability and upper bounds for spatial covariances have been obtained via Malliavin calculus and functional inequalities (see, e.g., [ 7 ] and the discussion in [ 10 ] ). Among the six fundamental KPZ initial geometries highlighted by Corwin [ 9 , Section 2.2 and Figure 2] , the flat profile is one of the most canonical: together with narrow-wedge and Brownian data, it gives a basic diffusive fixed point leading respectively to Airy 1 , Airy 2 , and stationary limits, and to GOE, GUE, and Baik–Rains fluctuations (see [ 9 ] and references therein). Historically, however, exact results for flat data have lagged behind those for narrow wedge, reflecting the greater difficulty of the flat regime: the Airy 2 covariance decay was first observed by Prähofer–Spohn [ 16 ] and rigorously proved by Widom [ 19 ] (see also [ 18 ] for higher-order terms), whereas the super-exponential Airy 1 decay was only identified much later in [ 3 ] . The present paper continues this timeline at the level of fixed-time KPZ covariances: Gu–Pu resolved the narrow-wedge case, and we now resolve the flat case. Motivated by [ 10 ] , we study how the fixed-time, large- |x| covariance depends on the initial profile ( 1.3 ). The main result of this paper is Theorem 1.1 below, which identifies the exact asymptotic of the covariance \Cov[h(t,x),h(t,0)] for flat initial data. Theorem 1.1 (Flat covariance asymptotic) . For flat initial data ( 1.3 ), for all t>0 , it holds that \displaystyle\Cov[h(t,x),h(t,0)] \displaystyle\sim\kappa(t)\int_{0}^{t}p_{2r}(x)\,\mathrm{d}r=\frac{2\kappa(t)}{\sqrt{\pi}}\,t^{3/2}\,|x|^{-2}\exp\!\left(-\frac{x^{2}}{4t}\right),\qquad|x|\to\infty, (1.4) \displaystyle\text{where}\quad\kappa(t) \displaystyle\coloneqq\Big(\mathbb{E}\!\left[Z(t,0)^{-1}\right]\Big)^{2}\quad\text{and}\quad p_{t}(x)=\frac{1}{\sqrt{2\pi t}}\exp\left(-\frac{x^{2}}{2t}\right). To our knowledge, Theorem 1.1 provides the first exact spatial covariance asymptotic for the KPZ equation under flat initial data. We refer to the asymptotic relation ( 1.4 ) as the flat-profile prediction. It is natural to view ( 1.4 ) together with the covariance decay of the universal KPZ scaling limits: the flat and narrow-wedge initial conditions lead (after the 1\!:\!2\!:\!3 KPZ scaling) to different Airy processes and, correspondingly, very different spatial decorrelation behaviors. We refer to [ 10 ] for a detailed discussion and further references. See also [ 19 , 18 ] for rigorous Airy 2 covariance asymptotics and [ 3 ] for correlation decay of the Airy 1 process. At the level of the KPZ scaling limits, the contrast between droplet/Airy 2 and flat/Airy 1 is particularly sharp. For positive functions f and g , we write f(x)\sim g(x) as x\to\infty if f(x)/g(x)\to 1 (in particular, \sim identifies the leading constant), while f(x)\asymp g(x) means two-sided bounds up to multiplicative constants for all sufficiently large x . With this convention, \displaystyle\begin{aligned} \Cov[\mathcal{A}_{2}(x),\mathcal{A}_{2}(0)]&=2x^{-2}+O(x^{-4}),&&x\to\infty&&\text{\cite[cite]{[\@@bibref{}{widom:04:on}{}{}]}},\\ \Cov[\mathcal{A}_{1}(x),\mathcal{A}_{1}(0)]&=\exp\!\left(-\left(\frac{4}{3}+o(1)\right)x^{3}\right),&&x\to\infty&&\text{\cite[cite]{[\@@bibref{}{basu.busani.ea:23:on}{}{}]}}.\end{aligned} (1.5) The polynomial decay in the Airy 2 case and the stretched-exponential decay in the Airy 1 case have a natural geometric interpretation in polymer/LPP models: for droplet data the two profiles are influenced by space–time regions with a macroscopic overlap (curvature forces geodesics/polymers to coalesce near a pinned endpoint), whereas in the flat case typical influential regions become essentially disjoint at large separation and correlations are driven by rare coalescence events. The fixed-time, large-separation covariances of the KPZ equation exhibit an analogous dichotomy. Writing h^{\mathrm{nw}} for the KPZ solution started from narrow wedge and h^{\mathrm{flat}} for the flat solution, Gu–Pu [ 10 ] *Theorem 1.1 and ( 1.4 ) together yield that for each fixed t>0 , \displaystyle\Cov\!\left[h^{\mathrm{nw}}(t,x),h^{\mathrm{nw}}(t,0)\right] \displaystyle\sim\frac{t}{|x|}, \displaystyle|x|\to\infty, (1.6) \displaystyle\Cov\!\left[h^{\mathrm{flat}}(t,x),h^{\mathrm{flat}}(t,0)\right] \displaystyle\sim\kappa(t)\int_{0}^{t}p_{2r}(x)\,\mathrm{d}r, \displaystyle|x|\to\infty. Thus the covariance decay changes from polynomial (narrow wedge) to Gaussian (flat), paralleling the Airy 2 versus Airy 1 contrast above. Heuristically, delta initial data pins every directed polymer at the origin, so h^{\mathrm{nw}}(t,x) and h^{\mathrm{nw}}(t,0) share noise near time zero regardless of separation, producing the slow |x|^{-1} tail; flat initial data lets the dominant paths originate from well-separated regions, and their noise overlap is mediated by heat-kernel diffusion, imprinting the Gaussian factor e^{-x^{2}/(4t)} . Remark 1.2 (Strategy and main difficulties) . The proof follows the Gu–Pu blueprint, with two additional flat-profile inputs. First, we use the Clark–Ocone formula to represent the covariance by an explicit deterministic kernel multiplied by a random boundary-layer factor. Second, we show a boundary-layer reduction: only times s\lesssim|x|^{-2+\varepsilon} contribute at leading order as |x|\to\infty . Third, on

Read the complete text