Study on Simulation of Foreshock Activity Properties before Strong Earthquake Using Heterogeneous Cellular Automata Models

Meng Li, Feng Yang, Tao ZhangInternational Journal of Geosciences 2014 · DOI

Abstract

This study utilizes 2-D random dynamic cellular automata models to simulate three degrees of heterogeneous fault structures, aiming to analyze the macroscopic evolution of seismic activity. The research demonstrates that fault heterogeneity significantly influences foreshock patterns and rupture processes. Specifically, moderately and less heterogeneous structures exhibit higher Accelerated Seismic Release (ASR) rates and more pronounced seismic gaps prior to main shocks. Furthermore, the distribution of stress drops for foreshock events narrows as a main shock approaches, indicating a transition toward mechanical homogeneity. These findings highlight the critical role of fault structure heterogeneity in the preparation processes of strong earthquakes.

Introduction: The "Black Box" of Earthquake Preparation

Predicting seismic events remains a formidable challenge in geosciences. While foreshocks—smaller tremors preceding a major quake—are frequently observed, distinguishing them from background seismic noise in real-time is difficult. This paper investigates how the physical complexity, or heterogeneity, of a fault line dictates these precursory patterns. By modeling different fault structures, researchers aim to identify specific seismic signals that reliably indicate an impending disaster.

Methodology: 2-D Cellular Automata Modeling

The researchers employed a 2-D Random Dynamic Cellular Automata (CA) model to simulate a fault plane as an 81x81 grid.

  • Modeling Heterogeneity: Using a 2D non-empty Cantor set, three distinct models were generated: Model A (most uniform), Model B (moderate heterogeneity), and Model C (highly heterogeneous/fractal).
  • Physical Parameters: Each cell possesses a specific static friction strength. Stress is added randomly and redistributes to neighbors upon rupture according to an inverse-square law.
  • Innovation: Unlike standard non-conservative models, this approach accounts for energy dissipation (e.g., heat or radiated waves), providing a more physically realistic simulation of earthquake dynamics.

Key Insights: Heterogeneity and Predictability

The simulation revealed that the degree of fault "patchiness" directly correlates with the type of seismic sequence produced:

  1. Model A (Uniform): Characterized by sudden, large failures (main-shock type) with minimal warning.
  2. Model B (Moderate): Produces classic foreshock-main-aftershock sequences, making it the most "predictable" model.
  3. Model C (Extreme): Results in earthquake swarms where constant small events mask the signal of a larger quake.
Evolution of stress states and earthquake magnitudes (M) over time (t) for three fault models: (a) Model A, (b) Model B, and (c) Model C.

As shown in the figure above, the stress state evolution and magnitude distributions vary significantly across the three models. A critical observation is the synchronization of stress drops: as a main shock nears, the range of stress drops narrows, suggesting the fault system acts with increasing mechanical unity as it reaches a tipping point.

Results: Statistical Performance and Seismic Gaps

The researchers evaluated the models using statistical standards to determine the reliability of precursory signals:

  • Detection Rates: Precursory activity was detected in 62% of cases for Model B (moderate heterogeneity), compared to 53% for Model A and only 36% for Model C. This confirms that moderate complexity provides the most reliable warning signs.
  • Seismic Gaps: These periods of relative quiet where stress accumulates are most prominent in smoother, less heterogeneous faults (Models A and B). In complex faults (Model C), chaotic stress redistribution prevents clear gap formation.
  • Magnitude vs. Complexity: While uniform faults (Model A) produced the highest individual magnitudes (up to 6.74), the most complex faults (Model C) generated a higher frequency of smaller events but lower peak magnitudes.

Conclusion and Future Directions

The study underscores that fault heterogeneity is a primary driver of earthquake preparation dynamics. However, the authors note limitations: the model is a 2D static simplification that does not yet account for 3D geometry, time-dependent friction (rate-and-state laws), or the actual velocity of crack propagation. Future research will focus on integrating these dynamic factors to create a more robust "virtual laboratory" for earthquake prediction.

Summary updated

Abstract

Three different degrees of heterogeneous fault models are simulated by using 2-D random dynamic cellular automata models for analyzing macroscopic behaviors of seismic activity evolution influenced by heterogeneity of fault structures. The results show that the heterogeneities of fault structures can influence evolution properties of the foreshock activity and rupture process, such as the mediate heterogeneous and less heterogeneous structures, which show relatively higher ASR rates and more significant seismic gaps before main shocks. Besides, stress drop distribution ranges of the foreshock events when approaching a main shock show more homogenous (narrower) than that of the foreshock events far from a main shock. So the heterogeneity of fault structures plays an important role in strong earthquake preparation processes.
BibTeX
@article{li2014study,
  author = {Meng Li and Feng Yang and Tao Zhang},
  title = {{Study on Simulation of Foreshock Activity Properties before Strong Earthquake Using Heterogeneous Cellular Automata Models}},
  journal = {International Journal of Geosciences},
  year = {2014},
  doi = {10.4236/ijg.2014.53028},
}
Figure