Study About The Development And Distribution Law And Quantitative Risk Assessment Of Landslide Disasters In Guizhou Province Based On Different Scales

Guizhou Province is a typical province of karst topography and geomorphology development in southwest China.Landslides occur frequently in Guizhou Province,which has caused immeasurable losses to people's lives and property.For example,in August 2017,landslides occurred in Dashujiao Formation of Pusha Community,Zhangjiawan Town,Nayong County,Guizhou Province,which caused serious casualties.The number of landslides similar to this one in the whole province is large,the types are various and the causes are complex.Landslide geological hazard risk assessment is the basis of landslide disaster prevention and mitigation.Through risk assessment,government departments can better help determine disaster mitigation measures?such as control engineering measures,monitoring and early warning?,so as to carry out scientific risk avoidance and control,and reduce the loss of personnel and property caused by landslides.On the basis of fully collecting and sorting out the relevant data of landslide geological hazards in Guizhou Province,this paper focuses on the development and distribution characteristics of landslide geological hazards in Guizhou Province,and analyses the key influencing factors of landslide with the theory of variable dimension fractal.On this basis,the susceptibility zoning of landslide geological hazards in Guizhou Province has been completed.Two typical objects in Guizhou Province are selected from the middle and high-risk areas.They are urban landslides concentrated area-Duyun Shabaobao office and mining landslides-Baimashan landslide in Xingren County.Quantitative risk assessment studies are carried out on the two scales of region and individual.The main achievements are as follows:?1?Development Distribution Law and Susceptibility Evaluation of Landslide Disasters in Guizhou Province.Firstly,it reveals the spatial and temporal distribution law of landslide disasters in Guizhou Province:according to the statistical investigation of landslides,landslide disasters in Guizhou Province have significant seasonality;used the spatial analysis method of GIS,the spatial distribution law of landslide disasters in Guizhou Province was analyzed.The results show that landslides do not exist in isolation,but have regular group characteristics;by calculating all landslides,medium and large landslides.The density of landslides in the west,southwest and northeast of Guizhou province is the most dense.Liupanshui city is the area adjacent to southwestern Guizhou and Bijie city,especially the area of Liupanshui is the relatively concentrated area of landslides and medium-large landslides.Secondly,it revealed the basic law of landslide disaster development in Guizhou Province:soil landslide is the main component of landslide,rock landslide is relatively small,and mostly carbonate rock and clastic rock interbedded,often has the characteristics of weak base and gentle inclination of rock stratum;from the length,width,area and volume of landslide,it analyses the relationship between landslide scale and frequency,and regression fitting ratio in the double logarithmic coordinate system.The conclusion showed that the key controlling factor of landslide scale trend in Guizhou Province is width.Then,the key control influencing factors of landslide are analyzed:the landslide in Guizhou Province is mainly controlled by seven key influencing factors,such as relative density and slope of coal mine.In this paper,the sensitivity and weight of each factor are calculated based on GIS and variable-dimension fractal theory.The results show that there are variable-dimension fractals in all seven influencing factors,which are second-order cumulative transformations.The order of factor importance sensitivity is coal mine relative.Density>Elevation>Slope>Distance from Fault>Stratum Lithology>Annual Average Rainfall>Distance from Water System.Finally,the fractal dimension of each factor is normalized to get the weight of each factor.Based on the grid unit,the susceptibility zoning of landslide geological hazards in Guizhou Province is completed by overlapping weighted information model in GIS.?2?Taking Duyun Shabaobao office as an example,the quantitative assessment of landslide risk at typical regional scale was studied.In this paper,eight index factors,including slope gradient,slope direction,section curvature,stratum lithology,elevation,distance from water system,distance from road and core density of landslide point,are selected to establish the susceptibility evaluation index system based on binary logistic regression model.According to the probability of scale and time of landslide,the landslide risk is evaluated quantitatively.The ROC curve is used to test the accuracy,AUC=0.803,and the prediction accuracy is good.Then,four representative indicators of buildings,population,land resources and traffic roads in the disaster-bearing body of the region are selected to be included in the vulnerability evaluation system.The spatial attribute database of disaster-bearing body is established by remote sensing interpretation and field investigation.The index values of each factor are quantified and normalized,and the vulnerability zoning of Shabaobao area is completed by overlapping weights and classifying them in GIS.Based on the empirical formula R=H*V,the risk index value and vulnerability index value are normalized and multiplied by GIS.Finally,the results of landslide risk assessment zoning in Shabaobao office area are obtained.?3?Taking Baimashan Landslide in Xingren County as an example,the quantitative risk assessment of typical single-scale landslides is studied.Firstly,qualitative analysis of landslide hazards is carried out.Combined with ground survey and high-density electrical prospecting,the present situation,causes and deformation characteristics of landslide are analyzed.The deformation trend and stability of landslide are analyzed by comparison of historical remote sensing images and InSAR technology.On this basis,the quantitative evaluation of landslide hazard is carried out.The mechanical parameters are determined by rock triaxial strength test and rock tensile strength test by splitting method,and the failure mode of landslide deformation and instability under continuous rainstorm is simulated by PFC^3D particle flow numerical simulation method.The dynamic road strength,moving time,accumulating body range,accumulating body thickness and moving speed of the sliding body are presented in three-dimensional manner.In the vulnerability evaluation,buildings and population are selected as vulnerability evaluation indicators,and three-dimensional modeling of building contour and UAV tilt photogrammetry is automatically extracted by eCognition software.Disaster-bearing body database is established from plane and elevation,and vulnerability evaluation of each evaluation factor is quantified.Finally,the risk and vulnerability of Baimashan single landslide are superimposed by using the spatial overlay function of GIS,and the result of risk assessment zoning of Baimashan single landslide is finally obtained.