Study of the creeping speed of Colluvial Deposits through Dimensional Analysis

Abstract

This document has the purpose of exposing the deduction of a law that allows predicting the speed of landslides on slopes formed by colluviums, affected by a high water table and water erosion, based on two predictive variables: the average rainfall and the average river flow that causes lateral undercut at the base. The methodology consisted of taking weekly to fortnightly topographic coordinates of points on the ground called landmarks, from which displacement and speed were calculated. The geological model has been determined from geophysical studies and soil tests. To develop the mathematical model, Buckingham's Pi Π Theorem has been used, relating the observed velocity and the characteristics of the soil with the predictive variables. The mathematical model obtained has given very close results compared to observed values ​​of speed. The main finding has been to establish the inverse proportionality between the crawling speed and the permeability of the colluvium. Its most important application is to identify precipitation and flow thresholds that would produce a catastrophic collapse on colluvial slopes, based on risk assessment models.