摘要
西藏地区地貌单元复杂、地质构造活跃,为该地区泥石流提供了良好的孕育环境,也对人类生命财产构成了极大威胁,开展泥石流易发性评价可为地区防灾减灾明确重点区域。以西藏自治区波密县−墨脱县为研究区域,利用Pearson卡方检验算法优选出高程、坡度、地层岩性、降雨量等12个对泥石流影响较高的因素作为评价指标,以研究区282个泥石流点和非泥石流点为样本数据库,基于ArcGIS平台利用信息量法和机器学习方法建立了4种易发性评价模型,并引入接受者操作特性(ROC)曲线和AUC(ROC曲线下方的积分面积)指标对泥石流易发性精度进行评估,得到了该地区的泥石流易发性评价分区。研究表明:(1)考虑不同维度泥石流类型主控因素不同,采用纬度和气温相融合的归一化系数作为泥石流易发性评价指标,在一定程度上消除了低海拔地区泥石流对温度的过度响应。(2)气温、距水系距离、距道路距离、地层岩性、高程是研究区泥石流发生的主控因素;植被覆盖率、地形湿度、坡度等因素也发挥着重要作用。(3)考虑泥石流灾害点和影响因子分级属性关系,对影响因子各分级属性赋分,作为输入特征进行训练,机器学习模型预测效果较好,平均AUC值为0.980,整体优于传统的信息量模型。(4)SVM模型的AUC值高达0.987,高易发区频率比率(FR)值为41.13且预测面积占比最小,具有在大尺度区域内进行高精度预测的能力。
Complex geomorphic units and active geological structures provide favorable conditions for debris flow in Xizang,which poses a great threats to human life and property.[Objective]The evaluation of debris flow susceptibility can identify key areas for disaster reduction and prevention in this region.[Methods]Taking Bomi and Motuo Counties of Xizang Autonomous Region as the study area,12 factors with high influence on debris flow,including elevation,slope,stratigraphic lithology and rainfall,were selected by Pearson Chi-square test algorithm as evaluation indexes.Data collected from 282 sits with and without debris flows in the study area were taken as the sample database.Based on ArcGIS platform,four susceptibility evaluation models were established by using Information Value Method and Machine Learning Method.The ROC curve and AUC index were introduced to evaluate the accuracy of debris flow susceptibility obtained from the proposed methods.[Results]A debris flow susceptibility map for the study area was obtained.[Conclusion]The results indicate that:(1)Considering different types of debris flows in different dimensions and controlling factors,the normalization coefficients of latitude and temperature are used as the evaluation index of debris flow susceptibility,which can eliminate the excessive responses of debris flow to temperature in low altitude areas to a certain extent.(2)Air temperature,distance from water system,distance from road,formation lithology and elevation are the main factors of debris flow occurrence in the study area;Factors such as vegetation coverage,terrain humidity,and slope also play an important role.(3)Considering the relationship between the disaster points of debris flows and the classification attributes of the impact factors,the classification attributes of the impact factors are assigned scores and trained as input features.The machine learning model performs well,and its average AUC is 0.980,which was better than the traditional information models.(4)The AUC of SVM model is as high as 0.987,and the FR value of the highly prone region is 41.13.The prediction area of high-risk regions takes up the smallest proportion,demonstrating superior highprecision prediction capability in large-scale regions.
作者
李群
徐红剑
杨金
王林康
孙靖宜
章广成
LI Qun;XU Hongjian;YANG Jin;WANG Linkang;SUN Jingyi;ZHANG Guangcheng(China Communications Second Highway Survey and Design Research Institute Co.,Ltd.,Wuhan 430058,China;Faculty of Engineering,China University of Geosciences(Wuhan),Wuhan 430074,China)
出处
《地质科技通报》
北大核心
2025年第4期316-329,共14页
Bulletin of Geological Science and Technology
基金
扎墨公路沿线地质灾害发育规律与工程影响评价研究。
关键词
卡方检验
评价指标优选
泥石流
易发性评价
高精度
信息量
机器学习
波密−墨脱地区
Chi-square test
indicator optimization
debris flow
susceptibility evaluation
high precision
information value
machine learning
Bomi-Motuo area
