Session Overview

To address the growing challenges of geohazards, we invite contributions that advance reliable and scalable machine learning/artificial intelligence (ML/AI) approaches for the detection, monitoring, and prediction of geohazards, such as earthquakes, tsunamis, volcanoes, landslides, and surface subsidence.

Focus Areas

We especially welcome works that:

• Multi-scale Sensing Integration: Integrate multi-scale sensing technologies (e.g., remote sensing and distributed fiber-optic sensing) with ML/AI to support both pre- and post-event assessment of geohazards as well as their cascading impacts in diverse environmental settings, such as urbanized, remote, post-disturbance, and cold landscapes.
• Interpretable ML/AI: Develop interpretable and knowledge-guided (e.g., physics-informed) ML/AI to reveal the driving factors and physical mechanisms.
• Model Robustness: Evaluate and improve model robustness in extreme and data-scarce scenarios through cross-region and cross-scenario model transfer, uncertainty quantification, and real-time data fusion.
• Open-Source Datasets: Develop open-source, multi-scale geohazard datasets to support the training and testing of foundation models toward reliable, scalable, real-world AI deployment.

Session Organizers