[Paper Review] Gap theorems for Kähler-Ricci solitons

Haozhao Li|ArXiv.org|Jun 30, 2009

Geometric Analysis and Curvature Flows13 references4 citations

TL;DR

This paper establishes two gap theorems for compact gradient shrinking Kähler-Ricci solitons with positive first Chern class: if the Ricci curvature is sufficiently close to the Kähler form (in L∞ norm), the metric must be Kähler-Einstein. The results provide quantitative thresholds based on the Futaki invariant, proving that non-Kähler-Einstein solitons cannot have Ricci curvature too close to the Kähler form.

ABSTRACT

In this paper, we prove that a gradient shrinking compact Kähler-Ricci soliton cannot have too large Ricci curvature unless it is Kähler-Einstein.

Motivation & Objective

• To establish quantitative gap theorems for compact gradient shrinking Kähler-Ricci solitons with positive Ricci curvature.
• To determine when a Kähler-Ricci soliton must be Kähler-Einstein based on the closeness of its Ricci curvature to the Kähler form.
• To provide explicit lower bounds on the gap between Ricci curvature and the Kähler form, in terms of the Futaki invariant.
• To support Tian’s conjecture on the limit of complex Monge-Ampère flow by showing that non-Kähler-Einstein solitons cannot arise from solutions close to t=1.

Proposed method

• Uses the Kähler-Ricci soliton equation $\mathrm{Ric}(\omega) - \omega = -\sqrt{-1}\partial\bar{\partial}u$ to relate the potential function $u$ to curvature and the Futaki invariant.
• Applies spectral theory on the Laplacian $\Delta_g$, deriving a lower bound $\lambda_1 \geq 1 - \epsilon$ for the first eigenvalue in terms of $\epsilon = \max_M |\mathrm{Ric}(\omega) - \omega|$.
• Employs integration by parts and $L^2$ estimates to derive the inequality $\epsilon^2 - (1 - \epsilon)f_X \geq 0$, leading to the gap threshold.
• For Theorem 1.3, normalizes $u$ with $\int_M u \, \omega^n = 0$ and derives a lower bound for $u$ using the soliton equation $\Delta_g u + u - |\nabla u|^2 = -f_X$.
• Uses diameter and scalar curvature bounds to control $|\nabla u|^2$, proving uniform upper bounds on scalar curvature $R$ in terms of $\lambda$ and $f_X$.
• Combines $L^2$ norm of $|\nabla\bar{\nabla}u|^2$ with curvature bounds to derive a contradiction when $\lambda \to 1$, forcing the metric to be Kähler-Einstein.

Experimental results

Research questions

• RQ1 Under what conditions on the Ricci curvature can a compact gradient shrinking Kähler-Ricci soliton fail to be Kähler-Einstein?
• RQ2 Can the Futaki invariant be used to quantify a 'gap' between Kähler-Ricci solitons and Kähler-Einstein metrics?
• RQ3 Does the complex Monge-Ampère flow on $c_1(M) > 0$ manifolds with no Kähler-Einstein metric necessarily produce a non-Kähler-Einstein soliton?
• RQ4 What is the optimal threshold $\epsilon$ such that $|\mathrm{Ric}(\omega) - \omega| < \epsilon$ forces the metric to be Kähler-Einstein?
• RQ5 How does the scalar curvature behave under the assumption $\mathrm{Ric}(\omega) \geq \lambda \omega$, and what does this imply for the soliton structure?

Key findings

• Theorem 1.1 establishes a sharp gap: if $|\mathrm{Ric}(\omega) - \omega| < \frac{-f_X + \sqrt{f_X^2 + 4f_X}}{2}$, then $\omega$ is Kähler-Einstein.
• The threshold in Theorem 1.1 is explicit and depends only on the Futaki invariant $f_X > 0$, showing that non-Kähler-Einstein solitons cannot have Ricci curvature too close to the Kähler form.
• Theorem 1.3 proves that if $\mathrm{Ric}(\omega) > (1 - \epsilon)\omega$ for sufficiently small $\epsilon > 0$ depending on $f_X$, then $\omega$ must be Kähler-Einstein.
• The scalar curvature $R$ of a Kähler-Ricci soliton is uniformly bounded above by a constant $\Lambda(\lambda, f_X)$ depending on $\lambda$ and $f_X$, with $\Lambda \to \text{finite}$ as $\lambda \to 1$.
• As $\lambda \to 1$, the $L^2$ norm of $|\nabla\bar{\nabla}u|^2$ tends to zero, contradicting the lower bound $\lambda f_X > 0$, forcing the metric to be Kähler-Einstein.
• The results imply that Tian’s conjecture on the limit of complex Monge-Ampère flow cannot produce non-Kähler-Einstein solitons if the flow approaches $t=1$ too closely.