宋渊, 江南, 张恩博, 江鸿彬. 基于证据权法的湖北省兴山县滑坡灾害易发性评价[J]. 地质与资源, 2023, 32(3): 352-359, 289. DOI: 10.13686/j.cnki.dzyzy.2023.03.012
    引用本文: 宋渊, 江南, 张恩博, 江鸿彬. 基于证据权法的湖北省兴山县滑坡灾害易发性评价[J]. 地质与资源, 2023, 32(3): 352-359, 289. DOI: 10.13686/j.cnki.dzyzy.2023.03.012
    SONG Yuan, JIANG Nan, ZHANG En-bo, JIANG Hong-bin. SUSCEPTIBILITY EVALUATION OF LANDSLIDE HAZARD IN XINGSHAN COUNTY OF HUBEI PROVINCE BASED ON WEIGHTS-OF-EVIDENCE METHOD[J]. Geology and Resources, 2023, 32(3): 352-359, 289. DOI: 10.13686/j.cnki.dzyzy.2023.03.012
    Citation: SONG Yuan, JIANG Nan, ZHANG En-bo, JIANG Hong-bin. SUSCEPTIBILITY EVALUATION OF LANDSLIDE HAZARD IN XINGSHAN COUNTY OF HUBEI PROVINCE BASED ON WEIGHTS-OF-EVIDENCE METHOD[J]. Geology and Resources, 2023, 32(3): 352-359, 289. DOI: 10.13686/j.cnki.dzyzy.2023.03.012

    基于证据权法的湖北省兴山县滑坡灾害易发性评价

    SUSCEPTIBILITY EVALUATION OF LANDSLIDE HAZARD IN XINGSHAN COUNTY OF HUBEI PROVINCE BASED ON WEIGHTS-OF-EVIDENCE METHOD

    • 摘要: 以三峡库区湖北省兴山县为例,综合分析历史滑坡数据及其孕灾因素,采用证据权法对县域进行滑坡地质灾害易发性评价.首先应用ArcGIS软件,提取高程、坡度、坡向、坡面曲率、断层、工程地质岩组、水系、植被指数评价因子图层,并按照评价因子各分级频率比划分证据层;其次,采用证据权法分别计算各证据层的权重及后验概率;然后将各证据层后验概率进行叠加,生成滑坡易发性分区图;最后,使用分位数回归法将研究区按滑坡易发程度分为极高、高、中、低易发区4类.采用成功率曲线法对证据权法评价精度进行验证,本次评价的精度为86.1%,表明评价效果较好.该评价方法及结果可为地方地质灾害"风险双控"提供技术依据.

       

      Abstract: Taking Xingshan County of Hubei Province in the Three Gorges reservoir region as an example, the paper comprehensively analyzes the historical landslide data and inducing factors, and evaluates the landslide susceptibility based on the weights-of-evidence method. First, ArcGIS is used to extract the evaluation factor layers including elevation, slope gradient, aspect, curvature, fault, engineering geological rock group, water system and vegetation index, and divide the evidence layer according to the frequency ratio of each evaluation factor. Secondly, the weights-of-evidence method is adopted to calculate the weight and posterior probability of each evidence layer. Then the posterior probability of each evidence layer is superimposed to generate the landslide susceptibility zoning map. Finally, the study area is divided with the quantile regression method into extremely high, high, medium and low risk areas in terms of landslide susceptibility degree. The evaluation accuracy of 86.1% for the method verified by receiver operating characteristic curve indicates that the evaluation effect is good, which can provide technical basis for risk control of local geological disasters.

       

    /

    返回文章
    返回