Abstract
Shenzhen is a megacity that emerged following China's reform and opening-up, having experienced rapid urbanization with frequent land transformation. Simultaneously influenced by a subtropical monsoon climate, the jurisdiction features abundant rainfall and dense vegetation cover, rendering geological hazard risks widely distributed across artificial slopes and natural slopes difficult to identify. First, using high-precision digital elevation models of the entire Shenzhen area obtained from airborne light detection and ranging (LiDAR) as the data source, over 3,500 typical markers for slope-type geological hazards were established through multi-source remote sensing identification. Second, comprehensively considering three major factors (eight factors) including topography, geological structure, and human engineering activities, and proposing a rainfall convergence factor based on the rainfall-induced disaster mechanism, the weights of evidence method was employed to complete hazard assessment of geological hazards under rainfall-induced conditions. Finally, a hazard threshold classification method for "key control points" was proposed within the actual context of individual hazards. The results indicate that airborne LiDAR technology can more extensively identify unstable slope bodies beneath vegetation layers, expanding the original computational sample database by approximately 330%, with the validation accuracy of the hazard assessment model reaching 0.903 on the receiver operating characteristic curve. The assessment results align with actual conditions, enabling effective evaluation of rainfall-induced slope failure probability, thereby forming a refined hazard assessment system for slope-type geological hazards applicable to densely vegetated urban areas.
Full Text
Hazard Evaluation of Slope Geological Disasters in Shenzhen Using Airborne LiDAR Data
DENG Bo¹, ZHANG Hui², BAI Jun¹, DONG Xiujun¹, JIN Dianqi², ZHANG Shaobiao²
¹State Key Laboratory of Geohazard Prevention and Geoenvironmental Protection, Chengdu University of Technology, Chengdu 610059, China
²Shenzhen Urban Public Safety and Technology Research Institute, Shenzhen 518046, China
Abstract
Shenzhen is a megacity that emerged following China's reform and opening-up, having experienced rapid urbanization and frequent land transformation. Influenced by a subtropical monsoon climate, the region receives abundant rainfall and has dense vegetation coverage, making it difficult to identify geological disaster risks widely distributed across artificial slopes and natural slopes.
Using high-precision digital elevation models of the entire Shenzhen area obtained from airborne Light Detection and Ranging (LiDAR) as the data source, we identified over 3,500 typical markers of slope geological disasters through multi-source remote sensing recognition. Considering three major factors (eight factors in total) including topography, geological structure, and human engineering activities, and proposing a rainfall convergence factor based on the rainfall-induced disaster mechanism, we completed a geological hazard risk assessment under rainfall-induced conditions using the weights of evidence method. Finally, we proposed a "key control point" hazard threshold classification method in the context of individual disaster backgrounds.
The results demonstrate that airborne LiDAR technology can more extensively identify unstable slope bodies beneath vegetation cover, expanding the original calculation sample database by approximately 330%. The receiver operating characteristic curve validation value of the hazard assessment model reached 0.903, and the evaluation results correspond with reality. The model can effectively assess the probability of slope instability induced by rainfall, forming a refined evaluation system for slope geological hazard risk applicable to densely vegetated urban areas.
Keywords: Shenzhen; Airborne LiDAR; Geological Hazards; Hazard Evaluation Model; Weights of Evidence Method