Deriving 3D displacement vectors from multi-temporal airborne laser scanning data for landslide activity analyses

Abstract

Information on geometries and kinematics of landslides are necessary to establish geological slope deformation models. We present two complementary geospatial methods to analyze landslide surface changes even in areas affected by strong surface pattern changes, making use of airborne laser scanning (ALS) data. An image correlation method based on shaded relief images with a uniformly diffuse lighting and a feature tracking based on terrain breaklines are applied on a data set of eight ALS flight campaigns analyzing an active deep-seated rockslide in the Eastern Alps (Austria). Both tracking methods are described in detail, including parameter assessment. Additionally, an accuracy assessment of the input data sets has been conducted. 3D vector displacement maps derived from image correlation are well suited for the study of landslides if only slight surface pattern changes occur. The smallest detectable displacements strongly depend on the accuracy of the ALS data and for image correlation results lie within the range of 0.24 and 0.75 m for this study. Displacement vectors derived by breakline tracking only allow to detect displacements greater than 2 m. However, in comparison to image correlation, breakline tracking is not limited to areas with slight surface pattern changes and allows us to detect displacements even in areas with strong surface pattern changes. For a comprehensive interpretation of landslide activity a combination of both methods, with consideration of additional supportive data such as elevation change images and orthoimages, is recommended.