Elliptic Harnack inequalities for mixed local and nonlocal $p$-energy form on metric measure spaces

This paper establishes weak and strong elliptic Harnack inequalities for mixed local and nonlocal $p$-energy forms on metric measure spaces by employing the De Giorgi--Nash--Moser method under axiomatic assumptions including the Poincaré and cutoff Sobolev inequalities.

The Gist

Imagine you are trying to understand the weather patterns in a very strange, complex city. This city isn't just made of smooth streets (like a normal city); it's a mix of smooth roads, sudden teleportation portals, and jagged, fractal-like alleyways.

In mathematics, this "city" is called a Metric Measure Space. The "weather" is a Harmonic Function—a value (like temperature or pressure) that has settled into a stable state, neither rising nor falling on its own.

The paper you shared is a guidebook for predicting how these "weather patterns" behave in this chaotic city. Specifically, it proves a rule called the Elliptic Harnack Inequality.

Here is the breakdown using simple analogies:

1. The Two Types of Movement (Local vs. Nonlocal)

In our city, things move in two ways:

• Local Movement (The Walk): Imagine a person walking down a street. They can only move to the next house over. This is smooth and continuous. In math, this is the "local" part (like the standard Laplacian).
• Nonlocal Movement (The Teleport): Imagine a person who can suddenly jump to a house miles away. This is "nonlocal." It's like a bird flying over the city or a particle jumping through space.

The paper studies systems where both happen at the same time. People are walking and teleporting simultaneously. This is the "Mixed" part of the title.

2. The Big Question: The Harnack Inequality

The Harnack Inequality is a rule that says: "If a temperature is stable in a neighborhood, it can't be freezing cold in one house and boiling hot in the house right next to it."

In a normal city (only walking), if a house is warm, the neighbors must be somewhat warm too. The ratio between the hottest and coldest spot in a small area is limited by a constant number.

The Problem: When you add "teleportation" (nonlocal jumps), this rule breaks.

• Why? Because a "teleporter" could bring a blast of heat from a distant volcano right into your living room, while your neighbor (who only walks) stays cold.
• The Fix: The authors had to invent a Modified Rule. They said, "Okay, the temperature in your house can be high, but only if there is a 'tail' of heat coming from far away." They added a special "Tail Term" to the equation to account for these distant jumps.

3. The Toolkit: How They Proved It

To prove this rule works for this messy, mixed city, the authors built a toolkit of three main tools (inequalities):

• The Poincaré Inequality (The Ruler): This measures how much the "temperature" fluctuates within a single room. It ensures the function isn't just chaotic noise; it has some structure.
• The Cutoff Sobolev Inequality (The Fence): Imagine you want to study a specific neighborhood. You need to build a fence around it to isolate it from the rest of the city. This tool tells you how much "energy" it costs to build that fence. If the fence is too expensive (too much energy), the math breaks. The authors proved that in this mixed city, the fence is always affordable enough.
• The Jump Conditions (The Teleport Rules): They made sure the "teleportation" isn't too wild. They assumed the jumps follow certain patterns (like "you can't jump too far too often"). This is the (TJ) and (UJS) conditions in the paper.

4. The Main Achievement

The paper proves that if your city has:

1. A consistent way of measuring volume (Volume Doubling),
2. A stable ruler (Poincaré),
3. Affordable fences (Cutoff Sobolev),
4. And reasonable teleport rules (Jump conditions)...

Then, the Modified Harnack Rule holds true.

This means that even in a world where things walk and teleport, we can still predict that the "temperature" (or any stable value) won't fluctuate wildly without a good reason (the "tail" from far away).

5. Why This Matters (The "So What?")

Before this paper, mathematicians had to solve these problems separately:

• One set of rules for smooth cities (walking only).
• Another set for teleporting cities (jumps only).
• And very few rules for the messy mix of both.

This paper unifies them. It's like discovering a single law of physics that explains both how a river flows (smooth) and how a flock of birds migrates (jumps), and how they interact when they meet.

The Result:

• Weak Harnack: We can bound the average temperature.
• Strong Harnack: We can bound the maximum temperature.
• Smoothness: Because the temperature can't jump wildly, the "weather map" is actually smooth and continuous (Hölder continuous), even in this chaotic city.

Summary Analogy

Imagine you are trying to predict the water level in a lake that is fed by a slow, steady river (local) and also by sudden, heavy rainstorms from miles away (nonlocal).

Old math said: "You can't predict the lake level because the rain is too random."
This paper says: "Actually, if we measure how much rain usually falls from a distance (the Tail), and we know the river flows steadily, we can predict the water level. We just need to add a 'rain factor' to our equation."

The authors have successfully written that equation for a vast, abstract class of mathematical worlds, proving that order can be found even in chaos.

Technical Summary

Here is a detailed technical summary of the paper "Elliptic Harnack inequalities for mixed local and nonlocal p-energy form on metric measure spaces" by Aobo Chen and Zhenyu Yu.

1. Problem Statement and Context

The paper addresses the regularity theory for mixed local and nonlocal $p$-energy forms on general metric measure spaces $(M, d, m)$.

• The Setting: The authors consider a space equipped with a $p$-energy form $(\mathcal{E}, \mathcal{F})$ that combines a strongly local part (analogous to the $p$-Laplacian, $-\Delta_p$) and a nonlocal part (analogous to the fractional $p$-Laplacian, $(-\Delta_p)^s$).
• The Challenge: While the elliptic Harnack inequality (EHI) is well-established for purely local operators (via the De Giorgi-Nash-Moser theory) and for purely nonlocal operators (under specific jump kernel bounds), the mixed case presents significant difficulties.
  • The lack of a bilinear structure (when $p \neq 2$) prevents the direct application of techniques used for Dirichlet forms ($p=2$).
  • Existing results for nonlocal equations often require the solution to be globally non-negative to satisfy the classical Harnack inequality. For locally non-negative solutions, "tail terms" (integrals over the exterior of the domain) must be included.
  • Previous literature often imposed strict upper bounds on the jump kernel (e.g., $J(x,y) \leq C d(x,y)^{-(d+sp)}$). The authors aim to relax these assumptions.

Goal: To establish both weak and strong elliptic Harnack inequalities for non-negative superharmonic and harmonic functions associated with these mixed forms, under a unified set of geometric and analytic assumptions.

2. Methodology

The authors employ a unified analytic framework based on the De Giorgi–Nash–Moser iteration technique, adapted for the nonlinear ($p \neq 2$) and mixed (local + nonlocal) setting.

Key Technical Components:

1. Axiomatic Definition: They introduce a rigorous definition of a mixed local and nonlocal $p$-energy form on a metric measure space. This involves decomposing the energy $\mathcal{E}$ into a local part $\mathcal{E}^{(L)}$ and a jump part $\mathcal{E}^{(J)}$ defined via a symmetric Radon measure $j$ on $M^2_{\text{od}}$.
2. Assumptions: The main results rely on the following conditions:
   • Geometric: Volume Doubling (VD) and Reverse Volume Doubling (RVD).
   • Analytic:
     • Poincaré Inequality (PI): Controls the oscillation of functions.
     • Cutoff Sobolev Inequality (CS): Relates the energy of cutoff functions to the $L^p$ norm.
     • Jump Kernel Conditions:
       • (TJ): A tail condition ensuring the jump measure doesn't grow too fast near the diagonal.
       • (UJS): Upper Jumping Smoothness, a mild condition replacing strict pointwise upper bounds on the jump kernel.
3. Core Estimates:
   • Caccioppoli Inequality: A reverse Poincaré inequality for subharmonic functions, controlling local energy by $L^p$ norms.
   • Growth Lemma: Establishes a relationship between the measure of super-level sets and the minimum value of the function.
   • Logarithmic Lemma: Proves that the logarithm of a superharmonic function has bounded mean oscillation (BMO), a crucial step for the weak Harnack inequality.
   • Tail Estimates: New estimates controlling the nonlocal "tail" of the solution using the supremum on a ball, derived using the (UJS) condition.
   • Mean Value Inequality (MV): Derived via De Giorgi iteration, linking the supremum of a harmonic function to its $L^p$ average.

3. Key Contributions and Results

Main Theorem 1 (Weak and Strong Harnack Inequalities)

The paper proves the following implications on a complete, volume-doubling, reverse volume-doubling metric measure space:

1. Weak Elliptic Harnack Inequality (wEH):
   $$(\text{CS}) + (\text{PI}) + (\text{TJ}) \implies (\text{wEH})$$
   This states that for a non-negative superharmonic function $u$, the $L^q$ average on a small ball is bounded by the essential infimum plus a nonlocal tail term.
   $$\left( \fint_{B_r} u^q \, dm \right)^{1/q} \leq C \left( \text{ess inf}_{B_r} u + W(B_r)^{\frac{1}{p-1}} T_{B_R, B_R}(u^-) \right)$$
2. Strong Elliptic Harnack Inequality (sEH):
   $$(\text{CS}) + (\text{PI}) + (\text{TJ}) + (\text{UJS}) \implies (\text{sEH})$$
   This states that for a non-negative harmonic function, the essential supremum is bounded by the essential infimum plus the tail term.
   $$\text{ess sup}_{B_r} u \leq C \left( \text{ess inf}_{B_r} u + W(B_r)^{\frac{1}{p-1}} T_{B_R, B_R}(u^-) \right)$$

Main Theorem 2 (Characterization of Cutoff Sobolev)

The authors establish a converse result, showing that the Cutoff Sobolev inequality (CS) can be derived from the Weak Harnack Inequality (wEH) combined with other standard conditions (VD, Faber-Krahn, capacity bounds). This provides a characterization of the functional inequalities in this setting.

Corollary: Hölder Continuity

As a consequence of the weak Harnack inequality, the authors deduce the Hölder continuity of harmonic functions. This confirms that solutions to these mixed equations are regular, extending known results from the linear ($p=2$) and pure cases.

4. Novelty and Advancements

• Relaxation of Jump Kernel Bounds: Unlike previous works (e.g., Chen-Kumagai-Wang) that required strict upper bounds on the jump kernel ($J \leq C d^{-d-sp}$), this paper only requires the (TJ) and (UJS) conditions.
  • Significance: The authors construct an example (Section 6.3, "Blow-up of a Cantor set") where the canonical upper bound fails, yet the Harnack inequalities still hold. This demonstrates the robustness of their axiomatic approach.
• Nonlinear Extension ($p \neq 2$): The theory is developed for general $p \in (1, \infty)$. The authors overcome the lack of bilinearity by developing new techniques for the energy of product inequalities and handling the nonlinearity in the tail estimates.
• Unified Framework: The results simultaneously recover known theorems for:
  • Purely local $p$-energy forms (where the tail term vanishes).
  • Purely nonlocal $p$-energy forms.
  • Mixed Dirichlet forms ($p=2$).

5. Examples

The paper validates the theory with three distinct examples:

1. Euclidean Space: The standard mixed operator $-\Delta_p + (-\Delta_p)^s$ on $\mathbb{R}^n$.
2. Ultrametric Space: A product of ultrametric spaces (e.g., $p$-adic numbers) with a specific nonlocal kernel, demonstrating applicability to fractal-like structures.
3. Blow-up of a Cantor Set: A purely nonlocal example on a fractal set where the jump kernel satisfies (TJ) and (UJS) but violates the standard upper bound condition, proving the necessity and generality of the new assumptions.

6. Significance

This work represents a major step forward in nonlinear potential theory on metric measure spaces. By establishing the Harnack inequality for mixed local-nonlocal $p$-energy forms under minimal and local assumptions, the authors provide a powerful tool for analyzing the regularity of solutions to complex equations arising in:

• Anomalous diffusion (where both local and long-range jumps occur).
• Random walks with long-range jumps.
• Mathematical finance and physics involving mixed operators.

The paper successfully bridges the gap between the classical local theory and the modern nonlocal theory, offering a unified perspective that is robust against the specific geometric and analytic properties of the underlying space.