Local Li-Yau's estimates on metric measure spaces

Hui-Chun Zhang, Xi-Ping Zhu

TL;DR

This paper extends the celebrated Li-Yau gradient estimates for the heat equation and Yau's gradient estimates for harmonic functions to a more general class of spaces called metric measure spaces satisfying the $RCD^*(K,N)$ condition. This is achieved by developing local versions of the Bochner formula and maximum principles adapted to these non-smooth settings, paving the way for deeper geometric analysis on these spaces.

What Problem Does This Paper Address?

The Li-Yau gradient estimate is a fundamental result in geometric analysis, providing a bound on the gradient of solutions to the heat equation on Riemannian manifolds. Similarly, Yau's gradient estimate offers crucial information about harmonic functions. These results have been extensively generalized to various settings, including weighted Riemannian manifolds and non-smooth spaces. However, extending these estimates to locally weak solutions of the heat equation on metric measure spaces satisfying the $RCD^*(K,N)$ condition, which are not necessarily smooth, presents a significant challenge. The standard techniques relying on smooth Bochner formulas and maximum principles are not directly applicable. This paper aims to bridge this gap by developing the necessary analytical tools to establish these important estimates in this more general and less smooth context.

Methods

The authors tackle this problem by building upon the existing framework for $RCD^*(K,N)$ spaces. The core of their approach involves:

  • Localizing the Bochner Formula: They adapt the global Bochner formula, which relates the Laplacian of a function to its derivatives and curvature, to a local setting. This requires defining and working with a "weak Laplacian" and carefully handling the properties of functions in Sobolev spaces on these metric measure spaces.
  • Developing Local Maximum Principles: Inspired by the classical maximum principle for harmonic functions, they establish versions of both standard and parabolic maximum principles that are suitable for weak solutions and the $RCD^*(K,N)$ setting. This involves extending concepts like "strict maximum" and utilizing the properties of the weak Laplacian.
  • Utilizing Steklov Averages: To handle the regularity issues of weak solutions, particularly the time derivative, the authors employ Steklov averages. This technique allows them to approximate the derivatives of weak solutions with smoother functions, enabling the application of the developed Bochner formulas and maximum principles.
  • Extending Kato's Inequality: A crucial step involves generalizing Kato's inequality, which relates the Laplacian of a function and its positive part, to the weighted measure spaces relevant to their analysis.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Bridging Smooth and Non-Smooth Analysis: The paper demonstrates how to adapt powerful tools from smooth geometric analysis (Bochner formula, maximum principles) to the more challenging realm of metric measure spaces with $RCD^*(K,N)$ conditions, which may not be smooth.
  • The Power of Localization: By developing local versions of key analytical inequalities, the authors can derive estimates that hold on specific regions of the space, which is crucial for understanding the behavior of solutions in more general settings.
  • Steklov Averages as a Regularization Tool: The effective use of Steklov averages highlights their utility in overcoming regularity limitations of weak solutions, making them a valuable technique for future work in this area.
  • Generalizing Fundamental Inequalities: The extension of Kato's inequality and the development of local maximum principles are significant contributions that broaden the applicability of these fundamental results.

Key Results

The paper establishes several key results:

  • Local Li-Yau's Gradient Estimate for Heat Equations: Theorem 1.4 provides a local gradient estimate for positive locally weak solutions of the heat equation on $RCD^*(K,N)$ metric measure spaces. This is a significant generalization of previous results.
  • Sharp Local Yau's Gradient Estimate for Harmonic Functions: Theorem 1.6 presents a sharp local Yau's gradient estimate for positive harmonic functions on $RCD^*(K,N)$ metric measure spaces, improving upon existing estimates.
  • Local Bochner Formula and Maximum Principles: The paper develops and proves local versions of the Bochner formula (Theorem 1.2) and maximum principles (Theorem 4.2 and 4.4) tailored for the $RCD^*(K,N)$ setting.

Limitations and Future Work

The paper focuses on metric measure spaces satisfying the $RCD^*(K,N)$ condition. While this is a broad class of spaces, it does not encompass all possible non-smooth spaces. The authors suggest that their methods could potentially be extended to other settings. Furthermore, the paper deals with locally weak solutions, and while they establish local estimates, further work could explore global estimates under appropriate conditions. The development of these local analytical tools opens doors for further investigations into the geometric analysis of $RCD^*(K,N)$ spaces.

Summary updated

Abstract

In this paper, we will study the (linear) geometric analysis on metric measure spaces. We will establish a local Li-Yau's estimate for weak solutions of heat equations and prove a sharp Yau's gradient gradient for harmonic functions on metric measure spaces, under the Riemannian curvature-dimension condition $RCD^*(K,N)$.
BibTeX
@inproceedings{zhang2016local,
  author = {Hui-Chun Zhang and Xi-Ping Zhu},
  title = {{Local Li-Yau's estimates on metric measure spaces}},
  booktitle = {Calculus of Variations and Partial Differential Equations},
  year = {2016},
  doi = {10.1007/s00526-016-1040-5},
}
Figure