基于逻辑回归模型和3S技术的思南县滑坡易发性评价  

Landslide susceptibility evaluation of Sinan County using logistics regression model and 3S technology

作　　者：胡涛[1] 樊鑫 王硕[3] 郭子正[1] 刘爱昌 黄发明 Hu Tao;Fan Xin;Wang Shuo;Guo Zizheng;Liu Aichang;Huang Faming(Faculty of Engineering,China University of Geosciences(Wuhan),Wuhan 430074,China;Geological Engineering Exploration Institute of General Administration of Geology and Mines of SINOPEC,Guiyang 550002,China;Land and Resources Bureau of Tongren City,Tongren Guizhou 554300,China;The Second Engineering Survey Institute of Guizhou Bureau of Geology and Mineral Exploration,Zunyi Guizhou 563000,China;School of Civil Engineering and Architecture,Nanchang University,Nanchang 330031,China)

机构地区：[1]中国地质大学(武汉)工程学院,武汉430074 [2]中石化地质矿山总局贵州地质勘查院,贵阳550002 [3]贵州省铜仁市国土资源局,贵州铜仁554300 [4]贵州省地矿局第二工程勘察院,贵州遵义563000 [5]南昌大学建筑工程学院,南昌330031

出　　处：《地质科技通报》2020年第2期113-121,共9页Bulletin of Geological Science and Technology

基　　金：国家自然科学基金项目(41807285);江西省自然科学基金项目(20192BAB216034);中国博士后基金项目(2019M652287);江西省博士后基金项目(2019KY08)。

摘　　要：区域滑坡易发性评价对滑坡灾害防治具有重要意义,贵州省思南县由于其特殊的自然地理和地质条件,受滑坡地质灾害的影响非常严重,因此,非常有必要对思南县的滑坡易发性进行评价。在滑坡编录的基础上,采用由RS、GIS和GPS组成的3S技术,获取了思南县的数字高程模型、坡度、坡向、剖面曲率、坡长、岩土类型、地表湿度指数、距离水系的距离、植被覆盖度和地表建筑物指数10个滑坡影响因子;再在频率比和相关性分析的基础上,利用逻辑回归模型对思南县的滑坡易发性进行了评价并绘制了易发性分布图。结果表明:利用逻辑回归模型预测思南县滑坡易发性的准确率(AUC值)达到0.797,较为准确地预测出了思南县滑坡分布规律;极高和高滑坡易发区主要分布在高程低于600 m、地表坡度较大且以软质岩类为主的区域;而极低和低滑坡易发区主要分布在高程较高、地表坡度较小且以硬质岩类为主的区域。The Sinan County of Guizhou Province,due to the specific and complex physical geography and geological conditions,is seriously affected by the landslide hazards.Hence,it is very necessary to conduct regional landslide susceptibility evaluation for landslide prediction and prevention in the area.This study uses 3S technology:remote sensing(RS),globe position system(GPS)and geographic information system(GIS),to evaluate landslide susceptibility based on the logistic regression(LR)model.The 3S technology is applied to obtain the landslide inventory,condition factors of landslides and other related basic data in Sinan County.About 308 landslides and ten affecting factors are acquired digital elevation model(DEM),slope,aspect,profile curvature,rock types,buffer of fracture lines,modified normalized difference water index(MNDWI),distance to river,normalized difference vegetation index(NDVI)and normalized difference building index(NDBI),using the 3S technology.Then based on the correlation analysis,LR model is used to calculate the landslide susceptibility indexes and map these indexes.Results show that,the area under the curve(AUC)of receiver operating characteristic curve(ROC)is 0.797 using LR model.The landslide distribution characteristics of Sinan County are accurately predicted by the LR model.In addition,the high and very high susceptible areas are mainly distributed in the areas where the DEM are higher than 600 m.In these areas,the slope are relatively great and the rocks are soft.The low and very low susceptible areas are mainly distributed in the areas where the DEM are high,the slopes are relatively low and the rocks are of hard rock class.

关 键 词：滑坡 易发性评价 逻辑回归模型 3S技术 频率比分析 思南县 

分 类 号：P642.22[天文地球—工程地质学]