滇东北地区滑坡易发性评价  

Landslide Susceptibility Assessment in Northeast Yunnan

作　　者：刘绍昌 白仙富[1] 储正伟 LIU Shaochang;BAI Xianfu;CHU Zhengwei(Yunnan Earthquake Agency,Kunming 650224,Yunnan,China;Yunnan Southern Geological Survey Engineering Co.,Ltd.,Dali 671000,Yunnan,China)

机构地区：[1]云南省地震局,云南昆明650224 [2]云南南方地勘工程有限公司,云南大理671000

出　　处：《地震研究》2025年第1期89-99,共11页Journal of Seismological Research

基　　金：云南省地震局震害防御处自立项目——基于CF与Logistic模型耦合的滇东北地震重点危险区滑坡易发性评价研究;云南省地震科技创新团队项目(CXTD202406).

摘　　要：在研究滇东北地震滑坡危险区滑坡致灾因子和孕灾环境的基础上,选取了地形、地质、其他共3大类12项评价因子,通过评价因子独立性检验,构建滑坡易发性评价指标体系。随机选取研究区70%的滑坡点作为滑坡易发性评价训练样本,运用GIS空间分析功能,分别采用确定性系数模型、二元逻辑回归模型以及二者组成的耦合模型进行滑坡易发性评价。选取未参与训练样本的30%滑坡点作为检验样本,使用受试者工作特征(ROC)曲线对以上3种模型的评价结果和精度进行验证。结果显示:高程、坡度、距道路距离、地质岩性、NDVI、年平均降雨量对滑坡易发性影响相对较大。耦合模型的AUC值为0.781,高于确定性系数模型的AUC值0.754和二元逻辑回归模型的AUC值0.748。耦合模型合理性和评价精度高于单一模型,且均符合模型检验要求。耦合模型揭示了各评价因子对滑坡易发性影响的差异性,能够较好地反映滑坡易发性影响因子之间相互权重大小,预测的研究区滑坡的空间分布与野外实际调查结果高度吻合。After studying the factors inducing landslides and the environment leading to landslides in the critical earthquake risk areas in northeast Yunnan Province,we selected 12 evaluation factors from 3 categories:“topography”,“geology”,and“others”.Then,through testing the independence of these factors,we constructed an evaluation index system of the landslide susceptibility.We randomly selected 70%of the landslide sites in the study area for the training samples of the landslide susceptibility evaluation.Using GIS spatial analysis function,we respectively used the Certainty Factor model,the Binary Logistic Regression model and the coupled model of these two models,to evaluate the landslide susceptibility in the study area.we used the rest 30%of the landslide sites for test samples,and adopted the Receiver Operating Characteristic curve to verify the accuracy of the evaluation results from the three models.We found that the evaluation factors like elevation,slope,distance from the road,geological lithology,NDVI,and average annual rainfall have relatively significant effects on the landslide susceptibility.AUC value of the coupled model is 0.781,higher than the one(0.754)of the Certainty Factor model and another one(0.748)of the Binary Logistic Regression model.The coupled model is more rational and accurate than the other two,and satisfies all the requirements of model test.The coupled model objectively revealed the difference of the influence of each evaluation factor,and well reflected the mutual weight of the influence factors of landslide susceptibility.In the study area,the predicted locations of the landslide by the coupled model is consistent with the locations determined through field survey.

关 键 词：滇东北 滑坡 易发性 耦合模型 

分 类 号：P315.94[天文地球—地震学]