Establishment of comprehensive remote sensing interpretation markers for geological hazards in Xuanwei City and evaluation of their application effectiveness

The study area is Xuanwei City, Qujing City, Yunnan Province, located in the southern section of the Wumeng Mountains on the central Yunnan-Guizhou Plateau. The overall landforms consist of deep-cut mid-mountain and low-mid-mountain terrain. The predominant tectonic line direction is NE–SW. The most widely exposed strata are the Permian Xuanwei Formation (P₂xn) and the Emeishan Formation (P₂β). The rock mass of the basement is soft and hard, with the rocks exhibiting weak weathering resistance and poor mechanical properties. Mining and quarrying represent the largest-scale human engineering activities in Xuanwei City, where coal mines are the most widely distributed, and goaf areas continue to expand. The main types of geological hazards include landslides, ground collapses, debris flows, and subsidence. These hazards are mainly small to medium in scale, with most currently stable or basically stable. To identify as many potential geological hazard points as possible and provide guidance for ground investigations, this study integrates conventional InSAR deformation data and optical images. By fully considering the development conditions and disaster-bearing elements of geological hazards, and establishing indicators related to susceptible environments and human activities, a set of comprehensive interpretation markers have been developed. This approach addresses the limitations of single-parameter interpretation methods, improves interpretation accuracy, and is applied effectively within the study area.

A total of 88 geological disaster points were identified and interpreted. To evaluate the effectiveness of these interpretation markers, an entropy weight-fuzzy evaluation method was developed to evaluate their quality. The comprehensive recognition rate was 65%, with 17% classified as basically correct and 18% as incorrect. Field verification confirmed that 61 points were correct, seven were basically correct, and 20 were incorrect. This corresponds to a recognition accuracy of 69%, the basic accuracy was 8%, and an incorrect rate of 23%. The actual verification results align with the quality evaluation results obtained with the use of entropy weight-fuzzy comprehensive method. Among the points, 35 were identified as existing geological hazards, and 33 were newly discovered geological hazards. Nineteen new points pose a threat to residents and are distributed across 13 townships, such as Baoshan, Dongshan, Longchang, and Tianba. A detailed list has been submitted to the Xuanwei Natural Resources Bureau to facilitate the implementation of appropriate prevention and control measures in the next phase.

The susceptible environment forms the basis for the occurrence of geological disasters, serving as a target area and aggregation for identifying potential hazards. Markers for landslide-prone environments primarily consider slope structures dominated by consequent and oblique slopes. Collapses mainly occur on slopes above 45°, particularly those with reverse, transverse, or oblique slope structures, as well as slopes exhibiting joint development. For the beds of debris flow gullies, a slope gradient exceeding the threshold of 30‰ and a catchment area larger than 0.18 km² are used as thresholds for interpreting markers of the water sources, while a total deposit volume greater than 5,000 m³ in the source area serves as the threshold for solid source markers. Karst collapses mainly involve carbonate strata, whereas goaf collapses are associated with coal-bearing strata and their direct overlying layers. A threshold of -20 mm·a^-1 is applied to InSAR deformation to detect goaf collapses, supplemented by the boundary of mining rights and identification of the markers of mining activities. Markers of human activities are obtained from land survey data and optical remote sensing images. For roads and railways, attention is given to whether muck yard stockpiles are located in ditches, the presence of buildings and structures downstream and nearby, and the potential formation of chain disasters. The results show that the comprehensive interpretation markers established in Xuanwei City exhibit good quality and effectively identify potential geological hazards.