Susceptibility Evaluation Of Landslide In Jiangzhou District,chongzuo City Based On Coupling Model

Jiangzhou District,located in the southwest of Guangxi and within Chongzuo City,features varying altitudes with higher elevations in the north and south,while the middle part is relatively lower.It is characterized by a rugged mountainous terrain with undulating peaks.The terrain and geomorphic conditions in Jiangzhou District are relatively complex,with a significant presence of geological disasters,primarily landslides and rockfalls.These disasters are well-developed,resulting in significant losses.In recent years,Jiangzhou District has experienced frequent occurrences of landslides and rockfalls,presenting a significant threat to the area.Currently,the absence of a landslide and rockfall susceptibility zoning map in the region poses substantial challenges to disaster prevention and mitigation efforts in the area.This paper focuses on the landslides and rockfalls in Jiangzhou District of Chongzuo City as the research subject.It analyzes the development characteristics,spatial patterns,and influencing factors of landslides and rockfalls based on data collection and field investigation.A correlation analysis was performed between the influencing factors of geological disasters in the study area and the distribution as well as the main geographical characteristics of disasters.Additionally,an evaluation index for the susceptibility of landslides and rockfalls was developed.Utilizing the frequency ratio model,logistic regression model,as well as the frequency ratio-logistic regression coupling model,in conjunction with Arc GIS,a grid calculation was conducted to evaluate the susceptibility.The evaluation results were categorized into five vulnerability zones: extremely high,high,medium,low,and extremely low vulnerable areas.Finally,the accuracy of the three models was tested,and the main achievements and conclusions are as follows:(1)The spatial distribution characteristics of the primary landslide and rockfall geological disasters in the study area were analyzed.In the study area,rock rockfalls and soil rockfalls are observed.However,the scale of rockfalls is small,and no large-scale rockfalls have occurred.They are primarily distributed in the middle and low mountains,as well as valleys,in the northern part of Jiangzhou District.In the southern region,they are found in low mountains and mausoleums.Additionally,peak cluster valleys in the central area also exhibit a scattered distribution.The landslides in the study area are primarily soil landslides,with a main distribution in the northern low mountains and valleys.A smaller number of landslides are also found in the southern low mountains and hills.The occurrence of landslides and rockfalls shows clear seasonality,with a higher frequency of disasters in June and July,which coincide with periods of frequent rainfall.On the other hand,there are fewer occurrences during December and January.(2)Statistical methods are utilized to analyze the collinearity among the data and the correlation between different evaluation factors and geological disasters.Eleven evaluation factors,including elevation,slope,slope aspect,TPI(Topographic Position Index),TWI(Topographic Wetness Index),average annual precipitation,distance to fault,rock formations,soil type,distance to water system,and land use,were selected.These factors were used to construct susceptibility evaluation indicators for landslides and rockfalls in Jiangzhou District,ensuring a rational assessment of susceptibility.(3)Frequency ratio models,logistic regression models,and frequency ratio-logistic regression coupling models were constructed to evaluate the susceptibility of landslides and rockfalls in this area.The areas with extremely high and high vulnerability,according to the three models,are mainly distributed in the northern,central,and southern regions of the study area.The accuracy of the three models can be evaluated using the receiver operating characteristic curve(ROC).The AUC values obtained from the area under the receiver operating characteristic(ROC)curve indicate that the frequency ratio model has an AUC of 0.72,the logistic regression model has an AUC of 0.75,and the frequency ratio-logistic regression coupling model has an AUC of 0.78.These values correspond to power generation rates of 72%,75%,and 78%,respectively.Precision decreases from highest to lowest in the following order: frequency ratiologistic regression coupling model,logistic regression model,and frequency ratio model.Overall,the coupling model is more suitable for evaluating the susceptibility of landslides in this area.