[Paper Review] Basic properties of the natural parametrization for the Schramm-Loewner evolution

Gregory F. Lawler, Mohammad A. Rezaei|arXiv (Cornell University)|Mar 14, 2012

Mathematical Dynamics and Fractals4 citations

TL;DR

This paper establishes foundational properties of the natural parametrization for Schramm-Loewner evolution (SLE$_\kappa$) with $\kappa < 8$, proving its existence, Hölder continuity relative to capacity parametrization, and domain independence. It provides up-to-constants bounds for the two-point Green’s function and shows convergence of expectations toward the Minkowski content, though the existence of the content itself remains unproven.

ABSTRACT

The natural paramterization or length for the Schramm-Loewner evolution (SLE{\\kappa}) is the candidate for the scaling limit of the length of discrete curves for \\kappa &lt; 8. We improve the proof of the existence of the parametrization and use this to establish some new results. In particular, we show that the natural parametrization is independent of domain and it is H\\"older continuous with respect to the capacity parametrization. We also give up-to-constants bounds for the two-point Green's function. Although we do not prove the conjecture that the natural length is given by the appropriate Minkowski content, we do prove that the corresponding expectations converge.

Motivation & Objective

• To establish the existence and fundamental properties of the natural parametrization for SLE$_\kappa$ with $\kappa < 8$, which is conjectured to be the scaling limit of discrete curve lengths.
• To show that the natural parametrization is independent of the domain and Hölder continuous with respect to the half-plane capacity parametrization.
• To derive up-to-constants bounds for the two-point Green’s function in SLE, a key quantity for understanding path geometry.
• To demonstrate that the expectations of the natural length converge to the Minkowski content, supporting the conjecture that the content defines the natural parametrization.

Proposed method

• Uses a refined proof of the existence of the natural parametrization via convergence of Minkowski content-like approximations.
• Applies the domain Markov property and conformal invariance to analyze the behavior of SLE paths under time and domain changes.
• Employs the two-sided radial SLE measure and local martingales involving $S_t = \sin \theta_t$ to control exit probabilities and derive asymptotic estimates.
• Uses Koebe's $1/4$-theorem and distortion theorems to control the image of disks under conformal maps $g_t$, enabling bounds on hitting probabilities.
• Applies strong Markov property and exponential decay estimates to compare hitting probabilities at different times, leading to exponential convergence in the Green’s function.
• Uses conformal invariance and distortion estimates to extend results from the unit disk to general simply connected domains with analytic boundary.

Experimental results

Research questions

• RQ1 Is the natural parametrization of SLE$_\kappa$ independent of the domain in which it is defined?
• RQ2 What is the regularity of the natural parametrization relative to the capacity parametrization?
• RQ3 Can sharp up-to-constants bounds be established for the two-point Green’s function of SLE$_\kappa$?
• RQ4 Do the expectations of the natural length converge to the $d$-dimensional Minkowski content, even if the content itself is not proven to exist?

Key findings

• The natural parametrization for SLE$_\kappa$ with $\kappa < 8$ exists and is independent of the choice of domain, a key property for universality.
• The natural parametrization is Hölder continuous with respect to the half-plane capacity parametrization, with exponent $\alpha > 0$ depending on $\kappa$.
• The two-point Green’s function satisfies up-to-constants bounds: $G_D(z;w_1,w_2) \asymp \epsilon^{2-d}$ as $\epsilon \to 0$, where $d = 1 + \min\{\kappa/8, 1\}$.
• The expectation of the natural length converges to the Minkowski content, though the existence of the content as a limit remains unproven.
• For the unit disk, the Green’s function satisfies $e^{t(2-d)} q(t,\theta) \to \hat{c}$ exponentially fast as $t \to \infty$, with rate $\alpha/2 > 0$.
• The asymptotic behavior of the hitting probability $q(t,\theta)$ is controlled via a local martingale and conformal invariance, leading to precise exponential convergence estimates.