A Cheng-type Eigenvalue-Comparison Theorem for the Hodge Laplacian

This paper establishes a uniform upper bound for the eigenvalues of the Hodge Laplacian on closed Riemannian manifolds with lower bounds on Ricci curvature and injectivity radius, as well as an upper bound on diameter, thereby extending Cheng-type eigenvalue comparison theorems to differential forms without requiring sectional curvature bounds.

The Gist

Imagine you have a musical instrument, but instead of strings or a drum, it's a complex, curved shape like a planet or a twisted piece of dough. In mathematics, this shape is called a Riemannian manifold. Just like a drumhead has a specific pitch when you hit it, this shape has "notes" it can naturally vibrate at. These notes are called eigenvalues.

The lower the note, the easier it is for the shape to vibrate in that way. The first note (the lowest one) is especially important because it tells us about the shape's overall "stiffness" or how easily it can wiggle.

For a long time, mathematicians knew how to predict the highest possible pitch for a simple drum (a sphere or a flat sheet) if they knew its size and how curved it was. A famous mathematician named S.-Y. Cheng figured out a rule for this in 1975: If you know how big the drum is and how curved it is, you can put a ceiling on how high the pitch can be.

However, Cheng's rule only worked for simple "drums" (functions). This paper tackles a much more complicated instrument: the Hodge Laplacian. Think of this not as a single drum, but as a drum that can vibrate in many different "directions" or "layers" at once (like a drum that can also twist and turn).

The Problem

Previous attempts to find a "pitch ceiling" for these complex drums required very strict rules. They needed to know the curvature of the shape in every single direction (Sectional Curvature). This is like saying, "I can only predict the pitch if I know exactly how every single tiny grain of sand on the drum is arranged." That's a lot of information to ask for!

The authors, Anusha Bhattacharya and Soma Maity, wanted to know: Can we predict the pitch ceiling if we only know the "average" curvature (Ricci curvature) and that the shape isn't too squished together?

The Solution: The "Pixelated" Approach

The authors came up with a clever way to solve this. Instead of trying to analyze the whole complex shape at once, they broke it down into tiny, manageable pieces.

1. The Grid of Balls: Imagine covering your complex shape with a grid of small, non-overlapping bubbles (geodesic balls).
2. The Harmonic Radius: They made sure these bubbles were small enough that the shape inside them looked almost flat and simple (like a flat sheet of paper). They call this the "Harmonic Radius." It's like saying, "If you zoom in close enough, the Earth looks flat."
3. The Comparison: They calculated the pitch for each tiny bubble. Because the bubbles are small and simple, they could use Cheng's old rules to find the pitch for each one.
4. Putting it Together: They proved that if you know the pitch of the loudest bubble in your grid, you know the ceiling for the pitch of the entire complex shape.

The Big Discovery

They proved that even with fewer rules (just average curvature and a minimum size for the bubbles), you can still put a strict ceiling on the pitch of these complex vibrations.

• The Analogy: Imagine you have a giant, wobbly jelly. You don't need to know the exact molecular structure of the jelly to know how fast it can wobble. You just need to know how big the jelly is and that it's not too runny (curvature) or too sticky (injectivity radius). This paper gives you the formula to calculate the maximum wobble speed based on those simple facts.

Why Does This Matter?

1. Simpler Rules: It removes the need for complex, hard-to-measure data. Now, mathematicians can make predictions about shapes using only the "average" curvature, which is much easier to find.
2. New Applications: They used this new rule to solve a specific problem about 1-forms (a specific type of vibration). This helps in understanding the "connection Laplacian," which is used in physics and geometry to describe how things flow or connect on a curved surface.
3. Infinite Shapes: They also showed that this logic works even for shapes that go on forever (non-compact manifolds), like an infinite plane, by looking at how the pitch behaves as you look at bigger and bigger chunks of it.

In a Nutshell

Think of this paper as a new, more flexible rulebook for predicting how a complex, curved object vibrates. The authors realized that you don't need a microscope to see every detail of the object; you just need to break it into small, flat pieces, measure those, and then use a clever mathematical trick to figure out the limits for the whole thing. It's a "Cheng-type" theorem, meaning it extends a classic rule to a much more difficult and interesting set of problems.

Technical Summary

Here is a detailed technical summary of the paper "A Cheng-Type Eigenvalue-Comparison Theorem for the Hodge Laplacian" by Anusha Bhattacharya and Soma Maity.

1. Problem Statement

The paper addresses the problem of establishing uniform upper bounds for the eigenvalues of the Hodge Laplacian ($\Delta = dd^* + d^*d$) acting on differential $p$-forms on closed Riemannian manifolds.

• Context: Classical results by S.-Y. Cheng (1975) provided upper bounds for the eigenvalues of the Laplace-Beltrami operator (acting on functions) based on dimension, diameter, and a lower bound on Ricci curvature.
• Gap: Previous extensions of Cheng's theorem to differential forms (by Dodziuk and Lott) required sectional curvature bounds, which are significantly stronger and more restrictive than Ricci curvature bounds.
• Objective: The authors aim to prove a Cheng-type comparison theorem for the Hodge Laplacian under the weaker assumption of a lower bound on Ricci curvature, provided the manifold also has a lower bound on the injectivity radius (or equivalently, the harmonic radius) and an upper bound on diameter.

2. Methodology

The authors employ a discretization approach combined with harmonic coordinate analysis and domain decomposition. The core strategy involves:

1. Harmonic Coordinates: Utilizing Anderson's results to establish the existence of harmonic coordinate charts of a uniform radius ($r_H$) on manifolds with bounded Ricci curvature and injectivity radius. In these coordinates, the metric tensor $g_{ij}$ and its derivatives satisfy specific $L^q$ bounds, allowing for local Euclidean-like estimates.
2. Local Dirichlet Estimates:
   • The authors construct specific test forms within small geodesic balls (radius $\epsilon < r_H$).
   • They use a specific $p$-form $\omega_0 = dy_1 \wedge \dots \wedge dy_p$ in harmonic coordinates.
   • By analyzing the action of the codifferential $d^*$ on this form, they show that $d^*\omega_0 = 0$ in harmonic coordinates.
   • This allows them to bound the Rayleigh quotient for the Hodge Laplacian on a ball $B$ by a constant multiple ($2^{2p+1}$) of the Rayleigh quotient for the scalar Laplacian on the same ball.
3. Domain Decomposition (Discretization):
   • The manifold is discretized using an $\epsilon$-net (a set of points $X$ such that balls of radius $2\epsilon$cover$M$and balls of radius$\epsilon$ are disjoint).
   • They utilize a min-max principle and a comparison lemma for quadratic forms to relate the global eigenvalues of the manifold to the maximum of the local Dirichlet eigenvalues on the discretization balls.
4. Geometric Partitioning:
   • To relate the discretization size to the global diameter $D$, they consider a minimizing geodesic of length $D$.
   • They partition this geodesic into segments of length $D/k$ and construct disjoint balls around these partition points.
   • This links the index $k$ of the eigenvalue to the radius of the balls used in the local estimates.

3. Key Contributions and Results

Main Theorem (Theorem 1.2)

The paper establishes a Cheng-type upper bound for the $k$-th eigenvalue $\lambda_{k,p}(M)$ of the Hodge Laplacian on $p$-forms.

• Assumptions: $(M, g)$ is a closed $n$-manifold with $\text{Ric}_g \geq (n-1)\xi$, diameter $D$, and harmonic radius $r_H$.
• Result:
  $$\lambda_{k,p}(M) \leq 2^{2p+1} \lambda_0^D \left( B_\xi \left( \frac{D}{2k} \right) \right)$$
  for $k \geq \frac{D}{2r_H}$, where $\lambda_0^D(B_\xi(r))$ is the first Dirichlet eigenvalue of the Laplacian on a ball of radius $r$ in the space of constant curvature $\xi$.
  • For small $k$ ($k \leq \frac{D}{2r_H}$), the bound is determined by the harmonic radius $r_H$.
• Significance: This removes the need for sectional curvature bounds, relying only on Ricci curvature and harmonic radius (which is controlled by Ricci curvature and injectivity radius).

Corollaries and Explicit Bounds

• Non-negative Ricci Curvature (Corollary 3.2): For $\text{Ric} \geq 0$, the authors derive explicit polynomial bounds in terms of $k, D, n,$ and $r_H$. For large $k$, the bound scales as $O(k^2/D^2)$.
• Negative Ricci Curvature (Corollary 3.3): Explicit bounds are provided for $\text{Ric} \geq -(n-1)\xi$, showing dependence on the curvature lower bound $\xi$.
• Volume-Based Bounds (Theorem 3.4): An upper bound is derived in terms of the volume $V$ of the manifold, specifically for $k > \frac{2^n V}{\alpha_{n-1} r_H^n}$.
• Connection Laplacian (Corollary 3.5): Using the Bochner-Weitzenböck formula ($\Delta = \nabla^*\nabla + \text{Ric}$), the authors derive an upper bound for the first non-zero eigenvalue of the connection Laplacian on 1-forms under non-negative Ricci curvature.

Non-Compact Manifolds (Section 4)

• The authors extend their results to complete non-compact manifolds by considering an exhaustion of the manifold by compact balls.
• They define the supremum of the $L^2$ spectrum, $\sigma_p(M)$, and prove:
  $$\sigma_p(M) \leq 2^{2p+1} \lambda_0^D(B_\xi(r_H))$$
  This provides a uniform bound on the spectral gap for non-compact manifolds with bounded Ricci curvature and injectivity radius.

4. Significance

1. Weakening Curvature Assumptions: The primary contribution is generalizing Cheng's eigenvalue comparison theorem from functions to differential forms while relaxing the curvature condition from sectional curvature to Ricci curvature. This is a major step in spectral geometry, as Ricci curvature is a more natural and weaker condition in many geometric analysis contexts.
2. Uniformity: The bounds are uniform for the class of manifolds defined by lower bounds on Ricci curvature, injectivity radius, and upper bounds on diameter. This is crucial for convergence problems and the study of limits of Riemannian manifolds.
3. Methodological Innovation: The paper successfully adapts the discretization method (previously used for scalar Laplacians) to the more complex setting of the Hodge Laplacian by leveraging the specific properties of harmonic coordinates to control the interaction between the metric and the exterior derivative/codifferential.
4. Applications: The results provide explicit estimates for the spectral gap of the connection Laplacian, which has implications for the study of harmonic forms, Hodge theory, and the topology of manifolds with non-negative Ricci curvature.

In summary, this paper successfully bridges the gap between scalar eigenvalue comparison theorems and those for differential forms, achieving this under the minimal geometric assumptions of Ricci curvature and injectivity radius bounds.