Parabolic gap theorems for the Yang-Mills energy

1. Problem Statement and Motivation

Context:
Classical Yang-Mills theory involves two primary equations on a vector or principal bundle over a Riemannian manifold:

1. Instantons: Solutions to the first-order anti-self-duality (ASD) equation $F_A^+ = 0$. These are absolute minimizers of the Yang-Mills energy functional $YM(A) = \frac{1}{2}\int_M |F_A|^2 dV$.
2. Yang-Mills Connections: Solutions to the second-order Euler-Lagrange equation $D_A^* F_A = 0$.

The Gap Problem:
A central question in gauge theory is determining when a Yang-Mills connection must be an instanton (or flat). Classical "gap theorems" (elliptic results) state that if the energy or curvature of a Yang-Mills connection is sufficiently small (below a specific threshold), the connection is necessarily an instanton.

• Limitation: These classical results require the connection to satisfy the Yang-Mills equation a priori. However, solving the Yang-Mills equation is notoriously difficult in dimensions $n \geq 4$ due to gauge freedom and the "bubbling phenomenon" (finite-time singularities).

The Goal:
The authors aim to broaden the scope of these gap theorems by removing the assumption that the initial connection is a Yang-Mills solution. Instead, they investigate the Yang-Mills flow (the gradient flow of the energy functional):
$$\frac{\partial A}{\partial t} = -D_A^* F_A$$
They seek to prove Parabolic Gap Theorems: statements asserting that if an arbitrary smooth connection has sufficiently small curvature (measured in specific norms), the Yang-Mills flow starting from it will exist for all time and converge smoothly to an instanton (or flat connection). This effectively establishes a deformation retraction from a neighborhood of connections onto the space of instantons.

2. Methodology

The paper employs a combination of parabolic PDE analysis, geometric blow-up arguments, and compactness theorems.

• Yang-Mills Flow Analysis: The authors utilize the flow equation to evolve arbitrary connections. A key challenge is that the flow is not strictly parabolic due to gauge invariance. They rely on the strategy of Råde (based on DeTurck's trick) to establish short-time existence and well-posedness in Sobolev spaces ($H^k$).
• $\epsilon$-Regularity and Blow-up: To prove long-time existence and convergence, the authors use a "Type-II blow-up" argument. They assume a singularity forms and rescale the solution to obtain a limit on $\mathbb{R}^4$ (or a sphere). By analyzing the limit, they derive contradictions using known properties of Yang-Mills connections on the sphere.
• Optimal Split $\epsilon$-Regularity: A core technical innovation is the proof of an optimal "split" $\epsilon$-regularity theorem for the self-dual curvature component ($F^+$) along the flow. This allows them to control the growth of curvature specifically in the directions that prevent instanton convergence.
• Lojasiewicz-Simon Inequality: To establish convergence rates and the deformation retraction property, the authors employ a variant of the Lojasiewicz-Simon inequality. This inequality relates the energy gap to the gradient of the energy functional, ensuring that solutions do not get "stuck" in non-minimal critical points but instead converge to the set of minimizers (instantons).
• Morrey Norms: The authors utilize Morrey norms ($M^{2,4}$) rather than just $L^2$ norms. These norms are scale-invariant and better suited for controlling the behavior of curvature in higher dimensions and on specific manifolds like quaternion-Kähler spaces.

3. Key Contributions and Results

The paper presents three main theorems corresponding to different geometric settings:

A. Deformation to Instantons on the 4-Sphere ($S^4$)

• Theorem 1.1: On an $SU(r)$-bundle over $S^4$ with charge $\kappa$, if the initial energy satisfies $YM(A_0) < 4\pi^2(|\kappa| + 2)$, the Yang-Mills flow exists for all time and converges smoothly to an instanton.
• Significance: This improves upon previous elliptic gap results (Gursky-Kelleher-Streets) by applying to any connection in the energy range, not just Yang-Mills ones.
• Application (Taubes's Theorem): The authors provide a streamlined proof of Taubes's theorem that the moduli space of instantons on $S^4$ is path-connected. By using the flow as a deformation retraction, they bypass the complex Lusternik-Schnirelman theory used in the original proof.

B. Deformation to Instantons on Quaternion-Kähler Manifolds

• Theorem 1.3: On a compact quaternion-Kähler manifold with positive scalar curvature, if a pseudo-holomorphic connection has a small $M^{2,4}$ Morrey norm for its self-dual curvature component ($F^+$), the flow exists globally and converges smoothly to an instanton ($F^+ = 0$).
• Significance: This generalizes the $S^4$ result to higher-dimensional manifolds with special holonomy ($Sp(m)Sp(1)$). It establishes that the space of instantons is a neighborhood deformation retract of the space of pseudo-holomorphic connections.

C. Deformation to Flat Connections (General Dimension)

• Theorem 1.4: On a compact Riemannian manifold of dimension $n \geq 4$, there exists a threshold $\delta_1$ such that if the $M^{2,4}$ norm of the full curvature is less than $\delta_1$, the flow converges smoothly to a flat connection.
• Corollary 1.5: A vector bundle admits a flat connection if and only if the infimum of the $M^{2,4}$ norm of curvature over all connections is zero.
• Corollary 1.6 (Naito's Result Recovery): The paper recovers Naito's result that for topologically non-trivial bundles, if the initial energy is sufficiently small, the flow must develop a finite-time singularity (blow-up). This highlights a dichotomy: small curvature implies convergence to flatness only if the bundle is topologically trivial; otherwise, the flow blows up.

4. Technical Highlights

1. Optimal Split Regularity (Theorem 3.3): The authors prove that for the self-dual part of the curvature, a bound of the form $\sup |F^+| \leq C\sqrt{\delta}$ holds, where $\delta$ is the smallness parameter. This is sharper than standard regularity estimates and is crucial for the convergence proofs.
2. Sobolev Distance Estimates (Appendix A): They establish precise estimates on the distance between solutions in Sobolev norms, which is necessary to prove that the flow defines a continuous map on the space of connections (not just gauge equivalence classes).
3. Lojasiewicz-Simon Inequality (Appendix B): The authors prove a gauge-invariant $H^{-1}$ Lojasiewicz-Simon inequality for the Yang-Mills functional. This is the engine that drives the exponential convergence of the flow to the critical set (instantons or flat connections).

5. Significance and Impact

• Unification of Elliptic and Parabolic Theory: The paper successfully bridges the gap between static (elliptic) gap theorems and dynamic (parabolic) flow behavior. It demonstrates that the "gap" phenomenon is robust enough to hold for the gradient flow starting from arbitrary data.
• Simplification of Topological Proofs: By replacing the intricate Lusternik-Schnirelman arguments in Taubes's work with the direct analytic tool of the Yang-Mills flow, the paper offers a more accessible and geometric proof of the path-connectedness of instanton moduli spaces.
• Higher-Dimensional Gauge Theory: The results on quaternion-Kähler manifolds and general Riemannian manifolds extend the understanding of gauge theory beyond the 4-dimensional setting, providing new tools for studying the topology of bundles in higher dimensions.
• Clarification of Blow-up vs. Convergence: The paper rigorously delineates the conditions under which the Yang-Mills flow converges to a smooth limit versus when it blows up, linking these behaviors directly to the topology of the underlying bundle and the size of the initial curvature.

In summary, Dayaprema and Waldron provide a powerful analytic framework that uses the Yang-Mills flow to deform spaces of connections onto their minimal subspaces, resolving long-standing questions about the topology of moduli spaces and the stability of gauge fields.