Quantifying landslide strain localization phenomena using tensor analysis of multi-temporal lidar data

Abstract A fundamental understanding of landslide evolution requires characterizing how deformation localizes within the sliding mass, as these non-homogeneous zones provide crucial insights into how destabilization initiates, failure surfaces develop, and the overall kinematic behavior evolves. While traditional analysis often assumes uniform movement, this study presents a methodology to quantify intricate patterns of surface deformation at a fine scale, allowing for the direct analysis of localization behavior. By applying strain tensor analysis to high-resolution displacement fields derived from multi-temporal Uncrewed Aerial Vehicle-Light Detection and Ranging (UAV-lidar) and Structure from Motion (SfM) surveys, we compute the divergence, gradient, and curl fields for two distinct landslides: one translational and one rotational. This approach quantifies volumetric changes, translational strain, and rotational components, revealing unique kinematic signatures for each landslide type. The translational slide is characterized by alternating expansion-contraction patterns along its dip-line, whereas the rotational slide exhibits clear, separate bands of head subsidence and toe expansion, coupled with non-uniform rotation along the strike. This detailed characterization of strain localization provides direct observational evidence of the fundamental mechanisms governing landslide behavior. It offers a more nuanced, mechanistic understanding that advances the interpretation of slope instability, providing a stronger physical basis for hazard assessment and risk management.