Landslide features identification and morphology investigation using high-resolution DEM derivatives

In the last decade, development in remote sensing techniques has opened new avenues for studying the evolution of landscapes dominated by mass wasting processes. Conventional methods including field reconnaissance are time-consuming and resource-intensive. Thus, it is worth taking advantage of the high-resolution digital elevation model (HRDEM) to identify landslide features remotely and investigate landslide morphology. This research proposes a new technique of landslide feature identification and morphology mapping using computer-aided methods to enhance the visual interpretation of HRDEM. These computer-aided methods involve deep exploration of topographic information provided by HRDEM. In addition to the HRDEM, nine diverse HRDEM derivatives were used to maximise the morphological information captured by HRDEM. To compact and to better understand the morphological information, original HRDEM derivatives were transformed into the principal component (PC) domain. Based on PC composition provided by three initial PCs, it was possible to identify morphological signatures of landslides and represent them as the detailed landslide surface morphology maps. The presented methodology serves as an alternative means of landslide characterisation. It permitted the evaluation of slope morphology and the ability to reassess recent and future landslide activity on a comparative basis.