Coupling Invasion and First Passage Percolation

Basic Information

• Paper ID: 2511.05618
• Title: Coupling Invasion and First Passage Percolation
• Author: Aldo Morelli
• Classification: math.PR (Probability Theory)
• Publication Date: November 6, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2511.05618

Abstract

It is well known that the continuity of the first-order phase transition in Bernoulli bond percolation on the integer lattice is equivalent to the probability of a vertex being invaded in invasion percolation tending to zero. This paper provides a coupling between invasion percolation and first passage percolation with logarithmically uniform passage times, yielding a new equivalent condition.

Research Background and Motivation

Core Problem

The core problem addressed in this research is establishing a connection between Invasion Percolation (IP) and First Passage Percolation (FPP), particularly through coupling methods to study the continuity problem of Bernoulli bond percolation at the critical threshold.

Importance of the Problem

1. Theoretical Significance: For integer lattices Zd with dimension 2 < d < 11, whether an infinite connected component exists at the critical threshold pc,d remains an important open problem
2. Methodological Value: By coupling different percolation models, techniques from first passage percolation can be leveraged to study properties of invasion percolation
3. Application Prospects: Percolation models have important applications in multiple fields including epidemic spreading, network security, and quantum magnetism

Limitations of Existing Methods

• Chayes, Chayes, and Newman (1987) established the equivalence between invasion set density and the existence of infinite connected components at the critical threshold
• However, directly studying properties of invasion percolation is difficult, lacking effective analytical tools

Research Motivation

The author aims to establish a coupling between invasion percolation and first passage percolation with logarithmically uniform distribution, enabling the use of mature techniques from first passage percolation to study invasion percolation, thereby providing a new approach to solving the continuity problem of critical percolation.

Core Contributions

1. Establishing New Coupling Relationship: First construction of a precise coupling between invasion percolation and first passage percolation with logarithmically uniform distribution
2. Providing Equivalent Conditions: Offering new equivalent conditions for continuity of Bernoulli bond percolation (Theorem 3.1 and Corollary 3.2)
3. Theoretical Guarantees: Proving that with appropriate parameter selection, the two models behave consistently in local regions with high probability
4. Numerical Verification: Providing simulation results for logarithmically uniform first passage percolation on Z², observing evidence of power-law behavior

Detailed Methodology

Task Definition

Given a d-dimensional integer lattice Zd, consider the coupling of the following three models:

• Bernoulli Bond Percolation: Each edge is independently included with probability p
• Invasion Percolation: Starting from the origin, at each step invade the adjacent uninvaded edge with minimum weight
• First Passage Percolation: Each edge is assigned a passage time, defining geodesic distance between vertices

Coupling Construction

On the same probability space Ω = (0,1^E, F, P), independently assign weights w(e) ~ U(0,1) to each edge e, then set:

• Invasion percolation uses original weights w(e)
• First passage percolation passage times are τK,e = e^(Kw(e)), where K > 0 is the coupling parameter

Key Definitions

For ball BR = {v ∈ Zd : d(0,v) ≤ R}, define:

• Containment Relation: IP contains (K,R) logarithmically uniform FPP if and only if all vertices v satisfying TK(0,v) < TK(0,∂BR) are invaded
• Coupling Parameter: K(R,ε) := log|ER|/δ(R,ε), where δ(R,ε) = (1-(1-ε)^(1/|ER|))/(|ER|-1)

Core Theorem

Theorem 3.1: For any ε > 0, r ≥ 0, there exists R₀ = R₀(ε,r) such that for all R ≥ R₀:

1. PIP contains (K,R) logarithmically uniform FPP on BR ≥ 1-ε
2. P(K,R) logarithmically uniform FPP contains IP on Br ≥ 1-ε

where K = K(R,ε/2) = O(ε⁻¹R^(4d)logR).

Corollary 3.2: $\inf_{x∈Z^d} \liminf_{R→∞} P[T_K(0,x) < T_K(0,∂B_R)] = 0 \iff θ_d(p_{c,d}) = 0$

Technical Innovations

Consistency of Ordering Relations

By introducing the event Tδ (any two edges in ER differ in weight by at least δ), proving that under this event the invasion order <IP and first passage order <FPP are completely consistent before reaching the boundary (Lemma 4.1).

Boundary Effect Control

Proving that for fixed r, when R is sufficiently large, the probability that vertices in Br are invaded after boundary vertices can be arbitrarily small (Lemma 4.2), ensuring consistency of behavior between the two models in local regions.

Parameter Selection Strategy

Cleverly choosing K = O(ε⁻¹R^(4d)logR), balancing the probability requirement for weight separation and the scale parameter of the exponential distribution.

Experimental Setup

Simulation Environment

• Dimension: Two-dimensional lattice Z²
• Boundary: Using ℓ¹ ball BR and asymmetric boundaries for testing
• Parameters: K = K(R,0.01) chosen according to Theorem 3.1
• Number of Trials: 10,000 independent trials for each configuration

Evaluation Metrics

For each vertex x ∈ BR, record the proportion P(x) of occurrences of the event TK(0,x) < TK(0,∂BR).

Implementation Tools

The author provides a GitHub repository: Log Uniform First Passage Percolation Simulation

Experimental Results

Main Findings

1. Geometric Shape: Despite the boundary being an ℓ¹ ball, probability contours exhibit circular characteristics, indicating some geometric renormalization effect
2. Power-Law Behavior: On the slice y=0, the probability distribution approximately follows the form 1-|x|^α(R), where:
   • For R=1000, α≈0.23, correlation coefficient r=0.998
   • According to Theorem 3.1, α(R)→0 as R→∞
3. Boundary Invariance: Using different boundary shapes (such as asymmetric boundary -x+|y|=100) yields similar circular contours

Numerical Results

• Figure 2(a) shows the probability distribution heatmap for R=100
• Figure 3(b) demonstrates high precision of power-law fitting (r=0.998)
• Results for different R values (100,200,500,1000) show consistent power-law trends

Related Work

Historical Development

1. Origins of Percolation Theory: Broadbent and Hammersley (1957) introduced Bernoulli bond percolation
2. Invasion Percolation: Wilkinson and Willemsen (1983) proposed it inspired by fluid flow in porous media
3. Theoretical Connections: Chayes, Chayes, and Newman (1987) established the equivalence between invasion percolation and Bernoulli percolation

Related Research

• Damron, Lam, and Wang (2015) applied invasion percolation to study properties of two-dimensional first passage percolation
• This work extends this approach, establishing a more general coupling relationship

Conclusions and Discussion

Main Conclusions

1. Successfully established a probabilistic coupling between invasion percolation and logarithmically uniform first passage percolation
2. Provided new equivalent conditions for determining critical continuity of Bernoulli percolation
3. Numerical experiments support theoretical predictions, observing interesting geometric and algebraic structures

Limitations

1. Dimensional Restrictions: Theoretical results hold for all dimensions, but numerical verification is only conducted in two dimensions
2. Parameter Dependence: The coupling parameter K grows rapidly with R and ε, potentially affecting numerical stability
3. Open Problems: Although new equivalent conditions are provided, the original critical continuity problem remains unsolved

Future Directions

1. Utilize limiting shape theory of first passage percolation to study geometric properties of invasion percolation
2. Explore coupling possibilities under other distributions
3. Study numerical behavior in high-dimensional cases

In-Depth Evaluation

Strengths

1. Theoretical Innovation: First precise coupling between IP and FPP, with novel and technically sophisticated methods
2. Rigorous Proof: Complete mathematical proofs with proper handling of technical details
3. Practical Value: Provides new tools for studying difficult percolation problems
4. Numerical Support: Simulation results consistent with theoretical predictions, enhancing credibility

Weaknesses

1. Complexity: Selection of coupling parameters is rather complex, potentially limiting practical applications
2. Computational Cost: Growth rate of K makes large-scale numerical experiments difficult
3. Application Scope: Currently mainly theoretical results; practical application prospects require further exploration

Impact

1. Academic Value: Provides new research tools for the probability theory and statistical physics communities
2. Methodological Contribution: Coupling techniques may inspire research on other stochastic processes
3. Open Problems: Opens new avenues for solving important open problems in percolation theory

Applicable Scenarios

1. Theoretical analysis of critical percolation phenomena
2. Modeling of network reliability and propagation processes
3. Mathematical analysis of phase transition phenomena in statistical physics

References

The paper cites classic literature in the field, including:

• Broadbent & Hammersley (1957): Foundational work in percolation theory
• Chayes, Chayes & Newman (1987): Theoretical foundation of invasion percolation
• Wilkinson & Willemsen (1983): Proposal of invasion percolation model
• As well as recent related research and application literature

This paper makes an important theoretical contribution to the field of probability theory, connecting two important stochastic processes through clever coupling construction, providing new mathematical tools for studying core problems in percolation theory. Although technically complex, its innovation and potential impact merit attention.