Landslides rarely occur as isolated events, nor can they be fully understood through a single methodological lens. As climate change and human activities alter environmental forcing, assessing landslide hazards requires capturing complex physical processes across diverse spatio-temporal scales, from micro-scale geomechanics to slope stability and catchment-wide hazard chains. At the same time, rapid advancements in data-driven methods, remote sensing, and machine learning / AI offer novel tools for spatial and temporal prediction, albeit often at the cost of physical explainability.

This session aims to bridge the gap between process-based physical understanding and data-driven prediction in landslide and (multi-)hazard assessments. We seek contributions that showcase how coupling empirical data, physics-informed statistical frameworks, and geomechanical models can yield robust, transparent hazard forecasting.

We welcome submissions addressing, but not limited to, the following topics:
- Bridging geomechanics-based slope stability with broad-scale susceptibility and hazard assessments.
- Integrating machine learning with physical process understanding, including explainable AI tools and physics-informed data-driven methods.
- Assessing process chains and cascading hazard interactions, covering landslide lifecycles with integrated methodologies, especially under non-stationary climate and environmental forcing.
- Mapping uncertainty propagation and information loss when scaling models from slope to catchment levels.