【作者】 王慧； 文海家； 胡东萍； 谢朋；

【Author】 WANG Hui;WEN Haijia;HU Dongping;XIE Peng;School of Civil Engineering,Chongqing University;Key Laboratory of New Technology for Construction of Cities in Mountain Area of the Ministry of Education,Chongqing University;

【机构】 重庆大学土木工程学院； 重庆大学山地城镇建设新技术教育部重点实验室；

【摘要】 本文以重庆市城口县作为研究区域,基于地理空间数据库,结合逻辑回归模型,开展山地环境地质灾害易发性县级区划研究。在研究过程中选取坡度、坡向、剖面曲率、微地貌、坡位、距水系距离、岩性、距断层距离、距道路距离、高程、归一化植被指数(NDVI)、顺逆向坡、年平均降雨量等13个因子作为诱发地质灾害的影响因子,应用GIS技术将各个因子专题图层栅格化,建立地理空间数据库。采用逻辑回归模型,在R软件中进行逻辑回归模型训练,得到各个影响因子的回归系数。建立地灾等级评价图,并将危险等级划分为极低、低、中、高、极高5种类型。对于研究结果运用受试者工作特征曲线(ROC曲线)进行合理性检验,位于ROC曲线下的面积AUC值为0.901,结果表明逻辑回归模型适用于重庆市城口县山地环境地质灾害易发性区划研究。更多还原

【Abstract】 The study area is Chengkou County in Chongqing.The main goal of this paper is to carry out study of geological hazard susceptibility on the basis of logistic regression model and geospatial database in mountain area.The process contains three major phases.Firstly,thirteen factors were selected as impact factor in total based on geospatial data.they were slope angle,slop aspect,curvature,micro-landform,slope position,distance from drainage,lithology,distance from faults,distance from highway,elevation,normalized difference vegetation index(NDVI),consequent-reverse slop and mean annual precipitation.Secondly,rasterizing every layer on the basis of GIS technology and establishing geospatial database.Thirdly,Sample training was accomplished by R software,the value of regression coefficients was determined as well as geological hazard susceptibility map could be obtained.Meanwhile,the susceptibility can be divided into five grades;very low,low,medium,high and very high.At last,the evaluation result was tested by ROC curve,the value of AUC is 0.901 which indicates logistic regression model is highly available to study of geological hazard susceptibility in Chengkou County.更多还原