Scaling Properties of Avalanche Activity in the Two-Dimensional Abelian Sandpile Model

Basic Information

• Paper ID: 2510.09631
• Title: Scaling Properties of Avalanche Activity in the Two-Dimensional Abelian Sandpile Model
• Author: Anubhav Ganguly (IISER Mohali)
• Classification: cond-mat.stat-mech nlin.CG
• Publication Date: September 28, 2025
• Paper Link: https://arxiv.org/abs/2510.09631

Abstract

This paper investigates the scaling properties of avalanche activity in the two-dimensional Abelian sandpile model. Unlike traditional studies of avalanche size distributions, the author analyzes the site activity distribution, which measures the frequency with which sites participate in avalanches as particles are added throughout the lattice. Through numerical simulations on systems up to L=160, averaged over 10⁴ configurations, the probability distribution P(A,L) of site activity is determined. The results show that P(A,L) follows the finite-size scaling form P(A,L)~L⁻²F(A/L²). For small values A≪L², the scaling function exhibits F(u)~u⁻¹/², corresponding to P(A)1/L; for large activity AO(L²), the distribution decays as F(u)~exp(-c₃u-c₄u²). The crossover between the two mechanisms occurs at A*~0.1L², marking a threshold between typical and highly excitable sites.

Research Background and Motivation

Problem Definition

The core problem addressed in this study is understanding the statistical properties of local site activity in the Abelian sandpile model (ASM). Traditional research has focused on the avalanche size distribution s(R) (the total size of avalanches initiated from site R), while this paper proposes studying the activity function A(R) (the total number of times site R participates in all possible avalanches).

Research Significance

1. Deeper Understanding of Self-Organized Criticality (SOC): The ASM is a canonical model of SOC; understanding its local activity patterns is crucial for comprehending critical phenomena
2. Practical Application Value: In real systems, typically only local activity can be observed, rather than global avalanche statistics
3. Theoretical Complementarity: Provides information complementary to traditional avalanche size statistics, revealing which regions serve as frequent channels for avalanche propagation

Limitations of Existing Methods

1. Traditional research primarily focuses on power-law distributions of avalanche sizes Prob(s)~s⁻τ
2. Lacks systematic analysis of the frequency with which local sites participate in avalanches
3. The activity distribution Prob(A,L) has not been systematically studied in the literature

Core Contributions

1. First Systematic Study of Site Activity Distribution: Defines and analyzes the probability distribution of the activity function A(R)=∑_{R'}N(R,R')
2. Discovery of Novel Scaling Form: Establishes the finite-size scaling relation P(A,L)~L⁻²F(A/L²)
3. Identification of Dual Scaling Mechanisms: Discovers power-law behavior F(u)~u⁻¹/² in the small activity regime and exponential decay F(u)~exp(-c₃u-c₄u²) in the large activity regime
4. Determination of Crossover Scale: Identifies A*~0.1L² as the boundary between the two mechanisms
5. Provision of Theoretical Framework: Derives scaling exponents α=β=2 through normalization and expectation value constraints

Methodology Details

Task Definition

Given a stable configuration C of the two-dimensional L×L Abelian sandpile model, define:

• Excitation Function: s(R) = size of the avalanche triggered by adding a particle at site R
• Activity Function: A(R) = ∑_{R'} N(R,R'), where N(R,R') is the number of times site R topples when a particle is added at site R'

The research objective is to analyze the probability distribution P(A,L) of the activity function A(R) and its scaling properties.

Model Architecture

Abelian Sandpile Model Definition

1. System Setup: L×L square lattice with open boundary conditions
2. State Variables: Each site (x,y) has height Δ_{x,y}∈{0,1,2,3} (stable state), Δ_{x,y}≥4 is unstable
3. Dynamical Rules:
   • Particle Addition: Δ_{x,y} → Δ_{x,y} + 1
   • Toppling Rule: Δ_{x,y} → Δ_{x,y} - 4, each of four neighbors receives one particle
   • Boundary particles are lost

Activity Function Calculation Procedure

1. Fix stable configuration C
2. For each site R':
   • Add a particle at R'
   • Record the number of times site R topples N(R,R')
   • Reset to configuration C
3. Calculate A(R) = ∑_{R'} N(R,R')
4. Compile the distribution of A(R) over all sites

Technical Innovations

Scaling Theory Derivation

By assuming the general scaling form P(A,L)~L⁻ᵅF(a/Lᵝ), utilizing:

1. Normalization Condition: ∑_a P(A,L) = 1 → α = β
2. Expectation Value Constraint: ⟨A⟩~L² → 2β - α = 2

Derive α = β = 2, yielding P(A,L)~L⁻²F(A/L²).

Dual Scaling Mechanisms

1. Small Activity Regime (u≪0.1): F(u)~u⁻¹/² → P(A)~1/L
2. Large Activity Regime (u≳0.1): F(u)~exp(-c₃u-c₄u²)
3. Unified Form: F(u) = (c₁u⁻⁰·⁵ + c₂)e⁻ᶜ³ᵘ⁻ᶜ⁴ᵘ²

Experimental Setup

Numerical Simulation Parameters

• System Sizes: L = 20 to 160
• Boundary Conditions: Open boundaries
• Statistical Samples: 10,000 configurations per size
• Computational Resources: MARIO cluster

Data Collection Procedure

1. System reaches critical state
2. For each configuration C, calculate A(R) for all sites
3. Compile empirical probability distribution P(A=a,L)
4. Perform multi-size scaling analysis

Evaluation Metrics

1. Scaling Exponents: Determined by fitting log-log plots
2. Data Collapse Quality: R² = 0.9839
3. Expectation Value Verification: ⟨A⟩ ≈ 0.085L² consistent with theoretical prediction

Experimental Results

Main Findings

Small Activity Scaling Behavior

• For fixed small values a, P(A=a,L)~1/L with slope -1±0.015
• Scaling function g(a) = LP(A,L) ~ a⁻⁰·³⁸⁵±⁰·⁰³
• Can be fitted to the form g(x) ≈ c₁x⁻¹/² + c₂

Unified Scaling Form

• Complete distribution obeys P(A,L)~L⁻²F(A/L²)
• Scaling function F(u) = (c₁u⁻⁰·⁵ + c₂)e⁻ᶜ³ᵘ⁻ᶜ⁴ᵘ²
• Parameter estimates: c₃~10, c₄=O(1)

Crossover Phenomena

• Crossover scale A*~0.1L²
• Below A*: power-law behavior dominates
• Above A*: exponential decay dominates

Special Configuration Analysis

For rare configurations C_ε with mass exceeding standard configurations by εL²:

• ⟨A⟩_C ~ εL³ (leading term)
• Fraction of such configurations ~exp(-ε²L²)
• Results in F(x~εL) ≳ exp(-x²) for large x

Data Collapse Verification

Figures 6 and 9 demonstrate excellent data collapse across multiple system sizes under the scaling form, validating the theoretical predictions.

Related Work

Theoretical Foundations of Abelian Sandpile Model

1. Bak-Tang-Wiesenfeld Model 1,2: Pioneering work on SOC
2. Majumdar-Dhar Mapping 9: Equivalence between ASM and q→0 Potts model
3. Conformal Field Theory Description: Central charge c=-2, dynamical exponent z=5/4

Avalanche Statistics Research

1. Traditional Studies: Focused on power-law distributions of avalanche sizes Prob(s)~s⁻τ
2. Deterministic vs. Stochastic Models: Manna model and BTW model belong to different universality classes 12-18
3. Multifractal Analysis 17,18: Precise calculation of critical exponents

Novelty of This Work

1. Observable Innovation: Shift from studying s(R) to studying A(R)
2. Scaling Theory: Establishment of complete scaling theory for activity distribution
3. Practical Value: Provides theoretical foundation for partially observed systems

Conclusions and Discussion

Main Conclusions

1. Scaling Form Established: P(A,L)~L⁻²F(A/L²) is the universal form for activity distribution
2. Dual Mechanism Identified: Boundary effects for small activity (~u⁻¹/²) and bulk effects for large activity (exponential decay)
3. Crossover Scale: A*~0.1L² marks the transition from boundary-dominated to bulk-dominated regime
4. Theoretical Consistency: Satisfies normalization and expectation value constraints

Physical Significance

• Local vs. Global: A(R) reveals local participation patterns, complementing global avalanche statistics
• Hotspot Identification: High-activity regions serve as frequent channels for avalanche propagation
• Critical Properties: Boundary effects produce scale-free contributions, but system average is dominated by large avalanches

Limitations

1. Finite-Size Effects: Maximum system size L=160 may be insufficient to completely eliminate finite-size effects
2. Boundary Condition Dependence: Choice of open boundary conditions may affect small activity behavior
3. Numerical Precision: Significant deviations at A=2 reflect combinatorial constraints of toppling sequences

Future Directions

1. Larger Systems: Extension to larger sizes to verify asymptotic behavior
2. Different Boundary Conditions: Investigation of behavior differences under periodic boundary conditions
3. Experimental Verification: Application of theory to local observations in real SOC systems
4. Analytical Theory: Development of rigorous analytical theory for activity distribution

In-Depth Evaluation

Strengths

1. Conceptual Innovation: Research perspective shift from traditional avalanche sizes to site activity has significant value
2. Theoretical Completeness: Rigorous derivation of scaling relations through normalization and expectation value constraints
3. Sufficient Numerical Verification: Excellent data collapse across multiple sizes validates theoretical predictions
4. High Practical Value: Provides theoretical foundation for partially observed systems
5. Clear Presentation: Logical structure and rigorous mathematical derivations

Weaknesses

1. System Size Limitation: Maximum size L=160 may be insufficient to fully verify asymptotic behavior
2. Boundary Effect Treatment: Separation of boundary and bulk effects requires deeper analysis
3. Lack of Analytical Theory: Primarily relies on numerical results, lacking rigorous analytical derivations
4. Universality Verification: Requires verification of similar behavior in other SOC models

Impact

1. Theoretical Contribution: Provides new observables and analytical framework for SOC theory
2. Methodological Value: Scaling analysis methods generalizable to other complex systems
3. Application Prospects: Potential applications in neuroscience, seismology, and other fields
4. Reproducibility: Detailed method descriptions facilitate reproduction and extension

Applicable Scenarios

1. Neural Networks: Analysis of local activity patterns in neuronal firing
2. Seismic Systems: Study of activity statistics from local strain sensors
3. Financial Markets: Analysis of local market participation patterns in fluctuations
4. Ecological Systems: Investigation of species participation distribution in ecological networks

References

This paper cites important literature in the SOC field, including:

• Pioneering work by Bak, Tang, and Wiesenfeld 1,2
• Theoretical contributions by Dhar 4,5
• Exact results by Majumdar and Dhar 9
• Recent experimental advances 22,23

Overall Assessment: This is an innovative work in Abelian sandpile model research that provides a new perspective for SOC theory through the introduction of site activity distribution. While there are some limitations regarding system size and analytical theory, its theoretical framework is complete and numerical verification is thorough, making it of significant value for understanding local critical phenomena.