Landslide Susceptibility Evaluation And Information Management System Development In Minxian County

Minxian County is located in the south of Gansu Province,with complex topography,significant tectonic movement,great differences in meteorological and hydrological conditions,fragile ecological environment,widespread landslide,which pose a serious threat to people's life and property safety and local social and economic development.Therefore,it has important application value and practical significance to carry out landslide research and susceptibility evaluation in Minxian County,and realize landslide data informatization and dynamic comprehensive management..Taking the landslide in Minxian County as the research object,this paper interprets the landslide by remote sensing image data,and analyzes the spatial distribution characteristics of landslide disaster in Minxian County.Combined with the factors of topography,geological structure,stratum lithology,meteorology,hydrology,surface coverage and human activities,it adopts the logistic regression information coupling model(LR-I)and radial basis function neural network information coupling model(RBFNN-I).Based on the above data,the landslide information management system of Minxian County was developed by programming.This study has reference significance for landslide evaluation,prevention and system construction.The main results are as follows:(1)Based on the multi-source and multi-temporal remote sensing data,the remote sensing interpretation marks of landslide in Minxian County are established,and the landslide are interpreted by using 3D Remote Sensing Image Technology.Combined with the field survey,549 landslide points are recorded.Through statistical analysis,the overall distribution density of landslide in Minxian County is higher in the north and lower in the south,and the spatial distribution has the characteristics of regional agglomeration.(2)The AUC values of LR-I and RBFNN-I models are 0.827 and 0.853 respectively,which indicates that the two coupled models can objectively and accurately evaluate the landslide susceptibility in Minxian County,and the evaluation accuracy of RBFNN-I model is higher than that of LR-I model.(3)The results show that the extremely high,high,medium and low risk areas of landslide account for 10.41%,20.00%,27.17% and 42.42% of the total area of Minxian County,respectively;the distance from fault,rainfall,distance from road and NDVI are the main controlling factors affecting the distribution of landslide in Minxian County.(4)Minxian landslide information management system integrates five functional modules,including system management,document management,data management,analysis and prediction and application tools,and has a good visual interface,and can realize the information and dynamic comprehensive management of landslide data,which has certain practical value.