Abstract

Understanding and quantifying the time evolution of landslides has always engaged researchers because of the consequences of such phenomena on the stability of buildings and infrastructure, and the loss of life. Consider, e.g., the catastrophic Vajont landslide in northern Italy in 1963 which caused great damage and the death of $1,917$ people. The scientific literature reports both mechanical and phenomenological approaches to analyzing landslide evolution. This paper aims to fill the gap between such approaches by introducing a geometric stability analysis of experimentally measured displacements trends. The proposed analysis organizes the experimental data of a given event into a dimensionless chart. The overall set of displacement data is partitioned into a sequence of activity stages associated with different triggering factors. This preliminary, but fundamental step, allows recognition of the common growth properties of different landslide displacements, independently of the volume of the main moving body, the material composition, and so on. The second step consists of a power-law regularization of the experimental data that allows the computing of time derivatives of the dimensionless cumulative displacements up to the third order (velocity, acceleration and second acceleration, or jerk). The approximating functions are used to understand and quantify the behavior of an experimentally monitored landslide event, by tracking its activity stages into a stability chart that accounts for five different regimes. The robustness of the proposed procedure is demonstrated through application to many well-documented case studies.

Keywords

Mathematical Subject Classification

Milestones

Authors