基于确定性系数模型与逻辑回归模型耦合的地质灾害易发性评价——以贵州省开阳县为例

Geohazard Susceptibility Assessment Based on Integrated Certainty Factor Model and Logistic Regression Model for Kaiyang,China

贵州省开阳县地质条件复杂,区内地质灾害频发。以该县为研究区,选取高程、坡度、地形起伏度、坡向、工程岩组、斜坡结构、断层、水系、归一化植被指数(normalized difference vegetation index,NDVI)等9个影响因子,基于GIS的栅格数据模型,分别采用确定性模型(certainty factor,CF)、Logistic回归模型以及两种模型耦合的方法进行地质灾害易发性评价,利用灾害点在各等级下的分布和成功率曲线(receiver operating characteristic,ROC)对三种模型的评价精度进行检验。结果表明:CF模型与Logistic回归模型评价结果总体上一致,但是耦合模型的评价分区结果更加合理,精度更高。因此,基于耦合模型的评价结果,将开阳县地质灾害易发性等级分为高易发区、中易发区、低易发区和极低易发区。这为县级区域地质灾害易发性评价提供了理论指导和技术参考。

Kaiyang County,Guizhou,SW China,has complex geological conditions and suffers from frequent geological disaster.To assess the local geohazard susceptibility,nine factors were selected and analyzed,including altitude,the slope gradient,relief amplitude,slope direction,the engineering rock group,fault,drainage,the normalized difference vegetation index(NDVI)etc,using GIS-based raster data model,the geohazard level was accessed in logistic regression model,the certainty factor model,and the combination model.The assessment accuracy of the three models was compared based on the distribution of disaster points at each level and the receiver operating characteristic(ROC)curve.Results show that the combined model is the most reasonable,consistent,and accurate.Therefore,in geohazard level,the study area are divided into four zones,ie,high,intermediate,low,and very low.The results provide a theoretical guidance and technical reference for the evaluation of regional geological hazards at a county-wise level.