Estimating the time of slope failure is a topic of great importance in the field of landslide risk mitigation. Within this framework, time of failure forecasting methods based on the inverse velocity, typically intended as the extrapolation of linear trend lines of the inverse of velocity with time, are widely known as tools for early warning of slopes displaying accelerating trends of deformation rate. Although nominally simple, their correct application is actually tricky as many factors can influence displacement data and eventually heavily reduce the accuracy of the predictions. Such disturbing elements can be classified as noise caused by instrumental precision and as noise representing the diverging of a natural behavior with respect to an ideal inverse velocity trend. Hence correctly preparing the dataset is a pivotal and critical task. The present teaching tool describes how to filter displacement data by presenting three different approaches and discussing the results of their application to three large slope failure case histories in Italy, in order to improve, in retrospect, the reliability of the failure-time predictions. Procedures to automatically setup alarm levels of slope failure occurrence are consequently proposed for supporting the definition of landslide emergency response plans.
TXT-tool 2.039-3.4 Methods to improve the reliability of time of slope failure predictions and to setup alarm levels based on the inverse velocity method
Di Traglia F.;Casagli N.
2018-01-01
Abstract
Estimating the time of slope failure is a topic of great importance in the field of landslide risk mitigation. Within this framework, time of failure forecasting methods based on the inverse velocity, typically intended as the extrapolation of linear trend lines of the inverse of velocity with time, are widely known as tools for early warning of slopes displaying accelerating trends of deformation rate. Although nominally simple, their correct application is actually tricky as many factors can influence displacement data and eventually heavily reduce the accuracy of the predictions. Such disturbing elements can be classified as noise caused by instrumental precision and as noise representing the diverging of a natural behavior with respect to an ideal inverse velocity trend. Hence correctly preparing the dataset is a pivotal and critical task. The present teaching tool describes how to filter displacement data by presenting three different approaches and discussing the results of their application to three large slope failure case histories in Italy, in order to improve, in retrospect, the reliability of the failure-time predictions. Procedures to automatically setup alarm levels of slope failure occurrence are consequently proposed for supporting the definition of landslide emergency response plans.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
