Analysis on the volume expansion effect and influencing factors on loess landslides: a case study of the Heifangtai tableland in the Chinese Loess Plateau.

Landslide scale measurement and hazard risk assessment are crucial, especially since landslide deposits often have an expansion effect, and can lead to more serious secondary hazard events. In this study, we obtained substantial high-resolution terrain and orthophoto data through multiple UAV photography and field surveys. Based on GIS spatial analysis and landslide geometric models, sliding and deposit data for 42 loess landslides in the Heifangtai tableland were calculated. The results show that sliding volume and sliding area have the highest correlation with the power-law distribution ($V_e = 1.204A^{1.1360}$ V e = 1.204 A 1.1360 ). The ratio of $H_{max}/L_{max}$ H max / L max increases with the increase of sliding volume and deposit volume of the landslides, with an average value of 0.306. The ratio of H/L has a higher correlation with deposit volume, and the fitting expression is $H_{max}/L_{max} = 0.187V_D^{-0.123}$ H max / L max = 0.187 V D − 0.123 . The VEC of 42 loess landslides ranged from 1.05 to 2.53, while the AVEC was about 1.503. The correlation analysis show that the VEC is positively correlated with AEC, H_mov, H_mov/L_max and L_mov/L_max but negatively correlated with D_ave. The essence of the volume expansion effect is that landslide fluidity increases, resulting in looser deposit distribution, larger pore scale in the deposit and lower overall compaction. Loess collapse in the Heifangtai tableland lack protection from the liquefaction of the sliding mass basement, resulting in large VEC values. The semi-ellipsoid model is a valuable approach for landslide volume assessment in the Chinese Loess Plateau, yielding an average error rate of restoring the sliding volume of landslides of 13.55%. This study provides a useful reference for accurate evaluations at the landslide scale. [ABSTRACT FROM AUTHOR]