Landslides threaten nearly 8% of the global population, and together with avalanches and glacial lake outburst floods, represent some of the most destructive mountain hazards worldwide. Forecasting these is essential to reducing human and economic losses, yet two fundamental challenges remain: identifying where and when initial failure will occur, and predicting runout once triggered.

In this talk, Dr Lorenzo Nava will present recent advances addressing both problems. First, he will show how satellite-based offset tracking can detect slope pre-failure motion and reveal incipient instability across mountain regions. He will then discuss how neural emulation of physical runout models enables scalable prediction of post-failure dynamics. Combined, these results connect the detection of incipient instability with rapid consequence modelling, bringing us closer to operational forecasting of cascading mountain hazards.

Sign up on Eventbrite.

This event is part of the King’s Institute for Artificial Intelligence’s AI Frontiers series.

Meet the speaker

Dr Lorenzo Nava is an AI+ Senior Fellow at King’s College London working at the intersection of artificial intelligence, Earth observation, and natural hazard science. His research investigates how complex hazard systems evolve and develops AI-driven approaches to monitor, forecast, and manage cascading mountain risks.

He serves as Chair of the Working Group on Educational Materials within the UN Global Initiative on Resilience to Natural Hazards through AI Solutions, where he coordinates international activities on capacity building for AI use in disaster risk reduction.

Before joining King’s, Lorenzo was a Research Associate at the University of Cambridge. He holds a PhD in Geosciences from the University of Padova.