The terrain and geological conditions of southern and southeastern Tibet are complex, the climate is harsh, and the traffic con ditions are extremely difficult. The long linear projects such as railways and highways involve a very wide range. Conventional survey methods are used for the survey of geological disasters such as mudslides due to low efficiency and high risks. It cannot be fully ascertained. This article uses satellite remote sensing, onventional aerial remote sensing, high-resolution drone airborne lidar (LiDAR) and high-precision oblique photography technology in the linear engineering at different design stages in the complex and dangerous mountainnous areas such as high altitudes and large elevation differences in southeastern Tibet. Research on the application of provenance identification technology. In the feasibility study stage, satellite remote sensing has obvious advantages due to its wide coverage. In the preliminary design stage, aerial remote sensing can meet the higher accuracy requirements. For areas with high vegetation coverage, UAV airborne LiDAR can remove vegetation to obtain a true elevation model of the ground. The provenance is effectively identified, and high resolution three-dimensional images can be obtained by the application of oblique photography technology, which can determine the dangerous rock after a strong earthquake, and provide an effective means for the identification of the seismic source. The formation of "satellite-conventional aviation-high-precision drones" and other multi-source, three-dimensional, comprehensive survey methods from macro-micro-details, can provide reference for the identification of debris flow provenance in similar areas.