With the expansion of observational and experimental datasets, advances in machine learning methods, and growing computational capabilities, machine learning is becoming an increasingly important component of Solid Earth geoscience. These approaches are opening new avenues for investigating physical and chemical processes across spatial and temporal scales, from the surface to the deep interior, on Earth and other terrestrial planetary bodies. 

This session welcomes contributions spanning a wide range of methods and applications, including data compilation and mining, statistical learning, classical and deep neural networks, explainable AI, generative models, and related approaches applied to geophysics, geodynamics, geochemistry, structural geology, volcanology, petrology, mineralogy, and mineral physics. We encourage both methodological developments tailored to geoscientific problems and application-focused studies that yield new insights into Solid Earth processes. Example topics include data mining of geochemical, mineralogical, or volcanological datasets; deep learning-based geophysical inversion; physics-informed emulators for geodynamics and landscape evolution; and machine-learning-assisted multiscale modeling.

• EP - Earth and Planetary Surface Processes
• S - Seismology
• T - Tectonophysics
• V - Volcanology, Geochemistry and Petrology