| Landslide is one of the most common geological disasters in mountainous areas of China.In recent years,China's economy has developed steadily and rapidly,and it has become the second largest economy in the world.With the rapid development of economy,many previously inaccessible areas have carried out corresponding infrastructure construction and various engineering activities in succession.Therefore,it is necessary to evaluate the natural disasters such as landslides in these areas.Under the influence of plate movement and the action of the rapidly uplifting,the elevation of the southeastern edge of the Tibet Plateau rapidly drops from 4,000 m to 1,500 m.Therefore,the region is rich in hydraulic resources.The upper reaches of the Jinsha River are located in this region.Due to the influence of the rapidly uplifting of the Tibet Plateau,the geological tectonic activity in the region is strong,the terrain is undulating,the climate is complex,and landslides and other disasters occur frequently.Therefore,the evaluation of landslide disaster in the upper reaches of Jinsha River is of great practical significance to ensure the safety of local people's property and the safe development of hydraulic resources.This paper studied the distribution characteristics,susceptibility and risk of landslide hazards in the upper reaches of Jinsha River in Xulong-Benzilan reach,and provides guidance for the safety construction of Xulong hydropower station and Benzilan hydropower station in the study area.The major accomplishments are as follows:(1)Based on the basic morphological characteristics of typical landslides,the spatial investigation method of landslides in the study area was proposed,based on the topographic and geomorphic features,interpretation by visible optical remote sensing,interpretation by InSAR technology and on-site investigation.A total of 61 landslides and unstable accumulation bodies were identified in the 50 km reach of the study area.Among them,there are 14 landslides along the Dingqu River,a tributary of Jinsha River,among which 5landslides are determined on the right bank and 9 landslides on the left bank.There are 47landslides along the Jisha River,among which 24 landslides are determined on the right bank and 23 landslides on the left bank.The total area of the landslide is about 38.85 km~2,and the total volume is about 1.96×10~9 m~3.The inventory map of the landslides was made.Combined with the geological conditions,the scale and genetic mechanism of the landslide are analyzed.The results show that the landslides in the study area is mainly middle-thick large-giant landslide(accumulation body).There are mainly five types of landslide genetic mechanisms in the study area,among which the ratio of flexural and tensile crack(39)and creep tensile crack(13)is the highest.Based on the analysis of the stratigraphic lithology and geological structure characteristics of the landslides,it is found that the landslides mainly occurred in the Jinsha River tectonic belt.The Jinsha River ophiolite suite,the symbolic product of this tectonic belt,is prone to landslides because of its fractured rock mass,near-vertical occurrence and plate-shaped rock formation.By comparing the dating data,it was determined that the landslides in the study area basically occurred in the period of rapid uplift of the Tibet Plateau since the late pleistocene,and the uplift rate was up to 5 mm/a.By comparing the results of landslide occurrence time and ice age division,it is found that the occurrence of landslides also has a certain relationship with ice age.(2)By summarizing the mapping units commonly used in landslide susceptibility mapping,slope units divided by hydrologic analysis method and curvature watershed method were selected as the mapping units in this study.Through qualitative and quantitative comparison of the two mapping units,it is found that the slope unit divided by the curvature watershed method can identify the horizontal surface and the inclined surface.Moreover,the area of the divided slope unit is more concentrated,and the shape is mostly between isosceles triangle and square,and there are few elongated units.The slope angle and slope aspect change inside the unit are smaller,and the interior is more uniform.(3)By analyzing the geological environment of the study area and combining with relevant studies,7 categories of lithologic index,topographic index,tectonic index,vegetation index,river index,rainfall index and seismic index were selected,and a total of 14sub-evaluation indexes were used as evaluation indexes for the landslide susceptibility mapping in the study area.The special geological characteristics of the study area:three-dimensional climate characteristics and rapid uplift characteristics are considered.The vertical distribution of rainfall in the study area was fitted by the empirical formula of rainfall in the mountain area,and the antagonistic action of internal and external forces in the unit was represented by the straler integral value.The correlation analysis and principal component extraction of 14 evaluation indexes were carried out by principal component analysis,and the evaluation index system of 14 pre-selected evaluation indexes was transformed into a new evaluation index system containing 7 principal components.(4)Logistics Regression model,random forest model and artificial neural network model were selected to optimize the landslide susceptibility model in the study area based on five-flod cross-validation,statistical parameters,Kappa coefficient and AUC value.The results of optimization show that the stochastic forest model is the best method for evaluating the susceptibility of landslide in the study area,whether the slope unit is divided by hydrologic analysis method or the slope unit is divided by curvature watershed method.(5)By comparing the evaluation results of the two slope units,it is found that the prediction accuracy of the landslide susceptibility mapping in the study area is not much different,and the prediction ability is basically the same.However,considering the dividing principle,dividing process,area distribution,unit shape,degree of uniformity within the unit and the workload of subsequent manual modification,it is suggested that the curvature watershed method should be given priority in the relevant evaluation and analysis.(6)The landslide susceptibility maps of the study area were analyzed,and the landslide susceptibility was divided into five grades:very low,low,moderate,high and very high.The evaluation results show that the landslide susceptibility map of the two slope units occupies more than 85%of the total landslide area in the very high and high landslide susceptibility.It shows that the landslide susceptibility map is reasonable.(7)Using two slope units,three evaluation indexes of building density,farmland density and road density were selected,and the vulnerability of landslide in the study area was evaluated with analytic hierarchy process(AHP).The evaluation results show that the study area is mainly characterized by extremely low-low vulnerability grade,which accounts for about 90%of the total area of the study area.The middle-high vulnerability grade area is mainly concentrated in the administrative villages on both sides of Jinsha River and Dingqu River.(8)Based on the results of landslide susceptibility mapping and landslide vulnerability mapping in the study area,the landslide risk in the study area was evaluated.The results showed that the area with low-very low landslide risk level accounted for more than 90%of the total area of the study area.The moderate-high risk level areas are mainly located in the administrative villages of Xulong township,Maoding village and Benzilan which are distributed from the upper reaches of Jinsha River to the lower reaches.Therefore,it is suggested to strengthen the monitoring and prevention of landslide hazards in these areas. |
PDF Full Text Request