Assessing loess landslide volume using high-precision UAV-derived DEM: A case study of the 15 March 2019 landslide in Zaoling Township, Xiangning County in North China

Abstract

Landslides pose significant hazards on China's Loess Plateau, potentially resulting in catastrophic consequences for human life and property. A prompt assessment of landslide disasters is crucial in providing scientific decision support for post-disaster relief efforts and serves as a fundamental basis for risk assessment of secondary disasters such as debris flow. On March 15, 2019, a loess landslide occurred in Zaoling Town, Xiangning County, Shanxi Province. This event led to the collapse of three buildings, resulting in 20 fatalities and 13 injuries. Subsequently, high-resolution orthophotos and an unmanned aerial vehicle-derived digital elevation model (UAV-DEM) of the landslide area were obtained through photogrammetry. By analyzing the shape of the exposed main scarp of the landslide in the 0.2 ​m UAV-DEM and the softened layer, this study successfully identified an unexposed underground sliding surface. Through further analysis of the shape of the exposed main scarp of the landslide in the 0.2 ​m UAV-DEM and the softened layer, this study accurately reconstructed the unexposed underground sliding surface. To reconstruct the pre-landslide digital elevation model (DEM), Google Earth images depicting the pre-landslide terrain features were utilized. The volume of each component of the landslide was then calculated using the pre-, post- DEM, and the sliding surface. The calculation results reveal that volume of the landslide is about 54,000 ​m³, the expansion rate was determined to be 13.9%.