摘要
通过确定性系数与地理探测器模型耦合的方法研究耦合模型对泥石流灾害易发性评估的影响。基于遥感判识和野外考察确定安宁河流域内的198条泥石流沟作为评估样本,选取坡度、地面峰值加速度、地形起伏度、地层岩性、植被覆盖度、植被类型、地形湿度指数以及地貌演化指数作为评价因子,耦合确定性系数和地理探测器模型,分别计算各评价因子分级取值和因子权重,进行安宁河流域泥石流易发性等级分区。研究结果表明:(1)地理探测器方法定量分析安宁河流域内泥石流发育的主要贡献因子为地面峰值加速度、地形起伏度和坡度。(2)泥石流极高易发区和高易发区分别占流域总面积的9.67%和12.02%,高易发区主要分布在安宁河流域两岸且呈条带状分布。(3)利用ROC曲线对泥石流易发性评估精度进行验证,耦合模型的AUC值达到0.824,相比单种模型CF模型的评估精度0.786提升了4.8%,具有更高的准确率,评价效果更优。
The influence of the coupling model of the certainty factor method and the geo-detector model on debris flow susceptibility was studied. The 198 debris flows in Anning River valley were selected as the research background based on the analysis of the remote sensing and field survey data.Eight influencing factors including slope grade,PGA,topographic relief,stratum lithology,vegetation coverage,vegetation type,topographic wetness index,and geomorphic evolution index were chosen as the evaluation factors. The coupling certainty factor and geo-detector model were used to calculate the grading value and factor weight of each evaluation factor respectively to obtain the zoning map of the debris flow susceptibility in Anning River valley. The results showed that:(1)The PGA,RA and slope were the main factors affecting the development of the debris flow in the Anning River valley.(2)The extremely high-prone areas and the high-prone areas of debris flow accounted for 9.67%and 12.02% of the total basin area respectively. The high-prone areas were mainly distributed on both banks of the Anning River basin in strips.(3)The debris flow susceptibility assessment accuracy was verified by the ROC curve,the AUC value of the presented model reached 0.824,which was 4.8%higher than that of the single model of 0.786. The proposed coupling model has a higher accuracy rate and better evaluation effect.
作者
杜宇琛
葛永刚
梁馨月
孙庆敏
陈攀
DU Yuchen;GE Yonggang;LIANG Xinyue;SUN Qingmin;CHEN Pan(Institute of Mountain Hazards and Environment,Chinese Academy of Sciences,Chengdu 610041,China;Key Laboratory of Mountain Surface Process and Hazard,Chinese Academy of Sciences,Chengdu 610041,China;University of Academy of Sciences,Beijing 100049,China;School of Engineering,Tibet University,Lhasa,850000)
出处
《防灾减灾工程学报》
CSCD
北大核心
2022年第4期664-673,共10页
Journal of Disaster Prevention and Mitigation Engineering
基金
国家自然科学基金重点项目(41792432)
中国科学院战略性先导科技专项(XDA23090203)资助。
关键词
泥石流易发性
耦合模型
地理探测器
确定性系数
安宁河流域
debris flow susceptibility
coupling model
geo detector
certainty factor method
Anning River valley
