Xiaolong Li,Qi S. Zhang
Abstract:In this paper we prove matrix Li-Yau-Hamilton estimates for positive solutions to the heat equation and the backward conjugate heat equation, both coupled with the Ricci flow. We then apply such estimates to establish the monotonicity of parabolic frequencies up to correction factors. As applications, we obtain some unique continuation results under the nonnegativity of sectional or complex sectional curvature.
What problem does this paper attempt to address?
This paper aims to solve the problem of matrix Li - Yau - Hamilton estimates related to the heat equation on Riemannian manifolds and its conjugate heat equation under Ricci flow. Specifically, the author attempts to extend Hamilton's matrix estimates for static metrics to the case of Ricci flow without relying on the parallel Ricci curvature condition, thereby making these estimates more applicable. In addition, the paper also explores the applications of these estimates in establishing the parabolic frequency monotonicity and unique continuation properties.
### Main contributions
1. **Matrix Li - Yau - Hamilton estimates**:
- The author proves the matrix Li - Yau - Hamilton estimates satisfied by the positive solution \( u \) under Ricci flow:
\[
\nabla_i \nabla_j \log u+\frac{\kappa}{1 - e^{-2\kappa t}} g_{ij}\geq0
\]
where \( \kappa>0 \) is a constant, and \( (M^n, g(t)) \) is a complete Ricci flow solution with non - negative sectional curvature and \( \text{Ric}\leq\kappa g \).
2. **Monotonicity of parabolic frequency**:
- Through the above matrix estimates, the author establishes the monotonicity of parabolic frequency, that is:
\[
e^{(n - 2)\kappa t(1 - e^{-2\kappa t})} F(t)
\]
is non - decreasing under the non - negative sectional curvature condition, where \( F(t) \) is defined as:
\[
F(t)=\frac{I'(t)}{I(t)}=\frac{2D(t)+S(t)}{I(t)}
\]
Here \( I(t) \), \( D(t) \), and \( S(t) \) respectively represent:
\[
I(t)=\int_M u^2(x, t) G(x, x_0, t)d\mu_{g(t)}
\]
\[
D(t)=\int_M|\nabla u(x, t)|^2 G(x, x_0, t)d\mu_{g(t)}
\]
\[
S(t)=\int_M u^2(x, t) R(x, t) G(x, x_0, t)d\mu_{g(t)}
\]
3. **Unique continuation result**:
- As an application, the author proves a unique continuation result, that is, if the positive solution \( u(x, t) \) vanishes to infinite order at a certain point \( (x_0, t_0) \), then \( u\equiv0 \).
### Methods and techniques
- **Hamilton's tensor maximum principle**: In the compact case, use Hamilton's tensor maximum principle to prove matrix estimates.
- **Li - Yau estimates**: In the non - compact case, use Li - Yau estimates and Harnack inequality to control the growth of solutions.
- **Integral bounds**: Through integration and partial integration techniques, obtain integral bounds for the first - and second - order derivatives of solutions.
- **Minimum eigenvalue**: Consider the minimum eigenvalue of the symmetric two - tensor \( tuZ_{ij} \) and use the maximum principle of the scalar heat equation.
### Applications
- **Geometric analysis**: These estimates and monotonicity results have important applications in studying the analytic, topological, and geometric properties of manifolds.
- **Unique continuation**: These results are helpful for studying the unique continuation properties of elliptic and parabolic equations.
- **Node set estimates**: These results can also be used to estimate the size of the node set of solutions.
In general, this paper provides new tools and methods for the heat equation under Ricci flow by extending and improving the existing matrix Li - Yau - Hamilton estimates, and has important theoretical and practical value.