On the earthquake predictability of fault interaction models

Area of Science:

• Geophysics
• Seismology
• Complex Systems

Background:

• Earthquake occurrence exhibits space-time clustering, deviating from random processes.
• Physics-based models, utilizing Coulomb stress changes, are commonly used to describe earthquake triggering.
• The predictive power of these models for future earthquakes remains largely unquantified.

Purpose of the Study:

• To evaluate the ex-ante predictability gain of physics-based models in earthquake occurrence.
• To investigate the effectiveness of Coulomb stress changes in improving earthquake forecasts.
• To identify conditions under which earthquake predictability can be enhanced.

Main Methods:

• Simulations of synthetic earthquake systems based on fault interaction physics.
• Analysis of earthquake triggering using Coulomb stress changes.
• Assessment of predictability gain from physics-based modeling versus intrinsic variability.

Main Results:

• Physics-based models, even those rooted in fault interaction physics, often yield limited improvements in earthquake predictability.
• Predictability gains are observed only when stress changes from nearby earthquakes significantly outweigh background stress and inherent system variability.
• The effectiveness of Coulomb stress models is highly dependent on the magnitude of stress perturbations relative to noise.

Conclusions:

• The predictive skill of Coulomb stress-based models for large earthquakes is often marginal.
• Significant earthquake predictability enhancement requires stress changes substantially larger than background fluctuations.
• Understanding the interplay between stress changes and intrinsic variability is crucial for improving earthquake forecasting.