Evaluation of landslide risk based on synchronization of nonlinear motions in observed data

Abstract

This work proposes an approach to evaluate the landslide risk from observed data by means of synchronization effects of nonlinear motions, forming a judgment on the linkage between two physical phenomena, slope failure and synchronization motion. Phase synchronization of spatial motion is an essential prerequisite for the landslide occurrence, and the connection between phase synchronization behaviors associated with the triggering factor at different points, as a definite mark, constitutes the sufficient criterion to forecast the coming of the landslide risk. Both of them can be obtained through phase synchronization analysis based on the method of empirical mode decomposition, Hilbert transform and order parameter description of phase synchronization. Direct action of the triggering factor can cause a generalized synchronization behavior, and otherwise the connection between different synchronization behaviors must be searched from three movement patterns as follows: damage spread up the slope, damage spread down the slope and damage spread up and down the slope. The first one shows orderly appearing of the peak phase values at different points up the slope, the second has a connection that the phase lock values vary in the same rhythm for a longer time period, and the last is the combination of the first two cases. Investigations of Vaiont landslide occurred on October 9, 1963, in Italy, and Xintan landslide occurred on June 12, 1985, in China, examined the effectiveness of the proposed approach, and also, as examples, provided a possible principle and effective measures to put it into practice for the early warning of landslide.