Computational Seismology and Physics-Informed AI

Computational Seismology and Physics-Informed AIVolume II: Rate-and-State Friction, Physics-Informed Neural Networks, and Hidden Stress InferenceThe most critical processes governing earthquakes are often invisible.Deep within the Earth's crust, faults evolve through frictional memory, stress transfer, fluid interactions, and nonlinear state transitions long before rupture occurs. While modern sensors generate unprecedented volumes of geophysical data, the true challenge lies in inferring the hidden physical states that control seismic behavior.Volume II explores the frontier where fault mechanics meets artificial intelligence.Combining Rate-and-State Friction theory, thermo-poroelastic processes, Coulomb stress transfer, Physics-Informed Neural Networks (PINNs), neural inversion, graph neural networks, and hybrid AI architectures, this volume presents a rigorous framework for understanding, modeling, and inferring the evolving condition of fault systems.Through mathematical derivations, computational workflows, real-world case studies, and emerging research directions, readers will discover how modern computational intelligence can be integrated with physical laws to transform seismic analysis and forecasting.Designed for researchers, engineers, geophysicists, data scientists, and graduate students, this volume serves as the bridge between classical earthquake physics and the next generation of intelligent Earth-system modeling.The future of earthquake science lies not only in observing the Earth—but in understanding the hidden states that govern its evolution.