A data-driven complex systems approach to early prediction of landslides

Abstract

Landslides are a common natural disaster that claims countless lives and causes huge devastation to infrastructure and the environment. The recent spate of landslides worldwide has prompted renewed calls for better forecasting methods which could boost the performance of early warning systems in real time. Although the variety, volume and precision of monitoring data have steadily increased, methods for analysing such data sets for landslide prediction have not kept pace with the rapid advances in complex systems data analytics and micromechanics of granular failure. Here we help close this gap by developing a new model to analyse kinematic data using complex networks. Like no other, our model incorporates lessons learned from micromechanics experiments on granular systems, with a focus on space-time variations and correlations in motion germane to the precursory dynamics of localised failure. We apply our model to ground-based radar data and predict where failure locates in a rock slope, spanning hundreds of meters, almost two weeks in advance. This is a first step in a broader effort to quantify the probability of a landslide occurring within a specified time based on data on kinematics and common triggers such as precipitation.