| Lanzhou city is located in the northeastern edge of the Qing Hai-Tibet plateau and the junction zone of the loess plateau.With strong modern tectonic activities and complex stratigraphic lithology,coupled with the increasing intensity and expansion of human engineering activities in recent years,geological disasters such as landslides and landslides in this area become more frequently.In this paper,based on the field survey,reasonable evaluation factors are selected and the secondary indexes of the selected evaluation factors are divided reasonably.On this basis,two cartographic units and three evaluation methods were used to carry out the vulnerability evaluation,and the advantages and disadvantages of these vulnerability evaluation results were compared respectively.The main conclusions are as follows:1.The field geological survey,the study area,a total of 52 development have a landslide,geological disasters,the study area is mainly distributed in the eastern of the quaternary loess,in clay stone and former Cambrian system Angle of biotite schist strata,the echelon along the erosion and accumulation valley plain and erosion accumulation loess hilly-gully region are distributed in the transition zone of ribbon.2.Through statistics and theoretical analysis,it is found that the eight influencing factors including elevation,slope,slope direction,slope type,fault,stratigraphic lithology,road and gully river have a great influence on the collapse and slip in the research area.Among the factors of gully river,the method of assessing the risk of occurrence according to different river channel and bank slope types was put forward for the first time,and the understanding that different river channel and bank slope types had different effects on the risk of occurrence of collapse and slip was summarized.3.Frequency ratio method based on the influence factor of secondary index partition,and deterministic coefficient method is used to analyse the factors affecting the liability of single factor analysis,the results show that the landslide in the study area for 1580 ? 1640 m in altitude,slope direction of the range of 0 ? 50 °,slope more than 40 °,slope type for convex type,450 m distance from the fault range,90 m distance from the road range in the formation of paleogene and Precambrian have a higher vulnerability.4.According to the characteristics of the study area,the evaluation results of the susceptibility of Terrain units and raster units were compared.Through analysis and statistics,it is found that although it is difficult for the Terrain unit to analyze the influence of slope direction factors on landslide,it can analyze the influence of slope height factors that are difficult for the grid unit to accurately calculate on landslide.5.Based on two different mapping units,binary logistic regression method,sensitivity index method and analytic hierarchy process(AHP)method were selected to evaluate the vulnerability,and a total of 6 vulnerability evaluation results were obtained.The study was divided into high,medium,low and very low risk areas by natural discontinuities.By comparing the common location of the distribution of the highly prone areas,it is found that when the slope in the study area meets the Terrain and geomorphic conditions,is close to the fault,and the stratigraphic lithology is poor,it is very easy to collapse and slip under the influence of other inducing factors.6.The ROC curve method was used to test and compare the evaluation results with the evaluation accuracy.The results showed that the AUC value obtained by all the evaluation methods of susceptibility was greater than 0.8,indicating that the selected evaluation methods of susceptibility could all play a good role in the evaluation of susceptibility.The evaluation accuracy of grid cell is higher than that of terrain cell.The difference between the sensitivity index method and the binary logistic regression method is small,but both of them are higher than the analytic hierarchy process(AHP)in the knowledge-driven method. |
PDF Full Text Request