| Plateau mountainous areas are characterized by frequent stratigraphic activities,complex geological structure,low resistance to ecological environment,and easy to occur landslide disasters.Lancang County is located in Yunnan Province,with typical plateau and mountain characteristics.Based on the data,this paper investigates the spatial distribution characteristics of landslide geological hazards in Lancang County.From an objective point of view,the three characteristics(quantity,scale and density),landslide components and their influencing factors of landslip in Lancang County are summarized.Based on the influencing factors,the evaluation indicators were selected and processed in grades,and correlation analysis was carried out.The indicators with strong independence were applied to the evaluation model.Through the test of rationality and accuracy,the accuracy of the information content model,logistic regression model and their coupling model was compared,and the mesh of different sizes was used to verify the accuracy of the model and select the most suitable mesh size,so as to achieve the optimization effect of the model.Finally,combined with the actual situation,the risk zoning results of the optimized model were discussed and analyzed,the landslide risk distribution area of the county was defined,and the specific suggestions were put forward.The main research contents and conclusions of this paper are as follows:(1)Explore the development characteristics,distribution rules and influencing factors of landslips in Lancang County.The development characteristics of landslides are discussed from three aspects(quantity,scale,density)and composition pairs.The distribution law is analyzed from two dimensions of time and space.The formation of landslip in Lancang County is closely related to many factors such as natural structure,climatic conditions and human intervention.(2)To establish a landslide risk evaluation index system suitable for Lancang County.With elevation,slope,aspect,stratigraphic rock group,distance from fault,distance from water system,distance from road,vegetation coverage rate and rainfall as evaluation factors,quantification and grading treatment were carried out,and importance and correlation analysis was carried out.The results show that the elevation,distance from fault,distance from water system,distance from road and rainfall of the evaluation factors are closely related to landslide hazard,and each evaluation index has little influence on each other and has high independence.All indexes can be used in the evaluation model.(3)The information content model,logistic regression model and the coupled model of information and logistic regression are used to study the landslide hazard in the study area,and the information content,frequency ratio and logistic regression coefficient of each evaluation factor in the study area can be calculated,and the influence degree of evaluation factors on landslide disaster can be analyzed.Among them,the factors that have the greatest influence on the information content and the logistic regression coupling model are consistent with the significant characteristics of the plateau and mountain areas,and there is a correlation between the plateau and mountain characteristics and the importance of the evaluation factors.The contribution of evaluation factors to landslide disaster under different models is obtained,and then the data is assigned to the grid.The landslide risk value obtained by Arc GIS software is obtained,so that the different risk zones are obtained.Through testing,the coupling model of information quantity and logistic regression is the most accurate.(4)Mesh is one of the main factors affecting the accuracy of the evaluation model,and different mesh sizes will affect the accuracy of the evaluation.First,it is verified that the accuracy of the coupling model is the highest among the three models under different grids.Then,the influence of the mesh size on the coupling model is studied.Finally,it is shown that the accuracy of the information content and logistic regression coupling model is the highest when the mesh size is 20 m,and this mesh size is the most suitable for landslide risk analysis in the plateau mountainous area.(5)The study area can be divided into five levels: extremely high landslide risk area,high landslide risk area,moderate landslide risk area,low landslide risk area and very low landslide risk area.The area of different areas,the number of landslide disaster points and the density of landslide disaster points were obtained by calculating statistics of the five grade areas,and the prevention and control suggestions were put forward according to the results. |
