High-precision and high-resolution digital terrain model data are crucial for the investigation and in-terpretation of terrain characteristics in large-scale landslide prone areas. However, dense vegetation in Taiwan’s moun-tainous areas hinders direct interpretation of surface features. Because of its ability to capture multiple returns and filter out vegetation, light detection and ranging (LiDAR) technology is commonly used in mountainous terrain surveys. Unmanned aerial vehicles, which feature high maneuverability, increased payload capacity, and allowance for low flight altitude, can be combined with position and orientation systems and lightweight LiDAR technology to efficiently and rapidly acquire spatial data. This technique is particularly suitable for surveying difficult-to-access mountainous areas. In this study, an unmanned aerial vehicle LiDAR system, called lair X3-H, was used to survey large-scale land-slide prone areas in Youyelin (Meishan Township) and Leye (Alishan Township), Chiayi County, Taiwan. The surveys were conducted at a flight altitude of 80 m above ground level, with a lateral overlap of >50% and a cross-flight plan. To evaluate the accuracy and error of the point cloud data, the flight line adjustments and point density were examined. In Leye for example, checkpoint verification of point cloud data revealed root mean square error values of 6.68 hori-zontally and 6.23 cm vertically. Comparisons of point cloud thickness with high-precision terrestrial LiDAR revealed a precision of approximately ±5 cm.
Key Words: UAV-LiDAR, digital elevation model, large-scale landslide, point cloud