NH3.16 | From detection to prediction: linking landslide causes, triggers, and outcomes
EDI
From detection to prediction: linking landslide causes, triggers, and outcomes
Co-organized by GM3/HS13
Convener: Lisa LunaECSECS | Co-conveners: Sansar Raj MeenaECSECS, Luca Piciullo, Minu Treesa AbrahamECSECS, Luca Ciabatta, Oriol Monserrat, Yaser Peiro

Effective landslide risk reduction and response efforts require reliable detection, informed process understanding, and accurate prediction. Advances in data-driven landslide detection are accelerating post-event mapping and leading to a growing availability of multi-temporal landslide inventories. These datasets, in turn, are allowing researchers to obtain a deeper understanding of the causes and triggers that influence landslide activity from hillslope to regional scales. For example, in combination with hydroclimatic models, re-analysis products, and meteorological observations, such inventories are enabling improved quantification of dynamic hydro-meteorological conditions that trigger weather-related landslides. Similar efforts are revealing indicators of co-seismic landslide hazard and underlying causes of slope instability. These insights are being integrated into data-driven, predictive models that can inform hazard assessments, increase situational awareness, and aid warning.

This session aims to spur future research advances and operational application development by bringing together a wide range of perspectives from geomorphology, hydrology, meteorology, remote sensing, data science and beyond. We will additionally explore how artificial intelligence (AI) and other data-driven approaches can enhance traditional methodologies, offering new insights for landslide detection, process understanding, and prediction.

Topics may include:
• Detecting and mapping landslide activity with remote sensing data and/or point source terrestrial data
• Linking trends and variability in landslide activity to hydro-meteorological, geological, morphological, or other conditions to improve process understanding
• Development and testing of new methods and approaches, including statistical, machine learning, and AI-based approaches, to support landslide hazard assessment, prediction, and early warning