摘 要: |
Length (L), width (W), and aspect ratio (epsilon, L/W) are elementary parameters widely adopted to characterize landslide geometry. Aspect ratio (epsilon) had been associated with the propagation of landslide along its movement path. However, current automatic measurement algorithms have not effectively considered the path of landslide movement, therefore only measure path-independent instead of path-dependent geometric parameters. This paper proposes a new approach for measuring path-dependent landslide L, W, and epsilon, in which path-dependent landslide L (L-pdep) is measured as the path distance from the landslide crown to its tip along the central line of landslide movement, and path-dependent landslide W (W-pdep) is measured as the weighted average lateral extent, across the central line of landslide movement. Specifically, W-pdep is calculated indirectly through dividing the landslide area by its length (A/L-pdep), and path-dependent landslide epsilon (epsilon(pdep)) can be calculated by L-pdep(2)/A. The central line of landslide movement, i.e., landslide profile, is a three-dimensional (3D) polyline generated by the freely available software ALPA (Automatic Landslide Profile Analysis), and all these path-dependent geometric parameters can be measured either in 3D space or in two-dimensional (2D) space (in the horizontal plane). A case study is presented showing that path-dependent parameters can be effectively measured by the proposed profile-based approach, and could be significantly different from path-independent parameters especially for long-narrow and tortuous landslides. This paper also practically introduces new path-dependent definitions of landslide L, W, and epsilon, which inspires a worthwhile revisit of nomenclature. An overview shows that significant divergencies exist between current definitions of landslide L, W, and epsilon. In order to moderate the raised ambiguities and confusions, this paper accordingly suggests a physically indicative inclusive nomenclature somehow embracing all current representative definitions, and then appeals for constituting type-specific nomenclatures for each landslide type. However, some open questions still exist. We therefore suggest relevant academic organizations to work further on type-specific nomenclatures and effective measurement approaches for landslide L, W, and epsilon. |