基于层次分析法-熵权法耦合模型的地质灾害易发性评价——以河北省迁西县为例

    Susceptibility assessment of geological hazard based on analytic hierarchy process-entropy evaluation method coupling model: A case study of Qianxi County in Hebei Province

    • 摘要: 迁西县地质条件复杂,人类改造自然活动频繁,地质灾害较发育,严重威胁群众生命财产安全。在分析迁西县地质灾害孕灾地质条件发育特征基础上,选取高程、坡度、坡向、起伏度、工程地质岩组、距构造距离、距道路距离、距河流距离8个评价指标构建地质灾害易发性评价指标体系,依据共线性诊断和相关性分析对指标间的独立性进行验证。运用信息量(information value,Ⅳ)法求得各评价指标信息量值,依托层次分析法(analytic hierarchy process,AHP)和熵权法(entropy evaluation method,EEM)分别确定各指标主、客观权重后,采用最小信息熵确定指标综合权重,进而建立Ⅳ模型、加权信息量(weighted information value,WIV)模型与AHP-EEM耦合模型,评价结果依据自然断点法将地质灾害易发性划分为非、低、中、高4个易发等级,并采取受试者操作特证(receiver operating characteristic, ROC)曲线开展精度验证。结果表明:AHP-EEM耦合模型精度最高,研究区高易发区、中易发区、低易发区和非易发区面积分别为270.85 km2、462.09 km2、486.13 km2、241.93 km2,占比分别为18.54%、31.63%33.27%、16.56%;AHP-EEW耦合模型AUC为0.875 6,评价结果与隐患点实际分布情况贴合度最高,符合本区域地质条件特征的易发性评价。评价结果可为本区域后期防灾减灾工作提供参考。

       

      Abstract: Qianxi County features complex geological conditions and frequent human-induced modifications to the natural environment, resulting in significant development of geological hazards that pose serious threats to people's life and property. Based on the analysis of the geological conditions contributing to hazard development in Qianxi County, the authors selected eight assessment indicators to construct a susceptibility assessment index system, including elevation, slope gradient, slope aspect, relief amplitude, engineering geological rock groups, distance from structural features, distance from roads, and distance from rivers. The independence of these indicators was validated through correlation analysis and collinearity diagnostics. The information value (Ⅳ) method was applied to calculate the information values for each assessment indicator, and subjective and objective weights were determined using the analytic hierarchy process (AHP) and entropy evaluation method (EEM), respectively. The comprehensive weights were then derived using minimum information entropy, and three models were established, that is, Ⅳ, WIV and AHP-EEW coupling model. The susceptibility assessment results were classified into four levels (non-susceptible, low, moderate, and high) using the natural breaks method, and then validated by receiver operating characteristic (ROC) curve. The results indicated that the AHP-EEM coupling model achieved the highest accuracy, and the high, moderate, low, and non-susceptible zones covered 270.85 km2 (accounting for 18.54%), 462.09 km2 (accounting for 31.63%), 486.13 km2 (accounting for 33.27%), and 241.93 km2 (accounting for 16.56%), respectively. The AUC value of AHP-EEM coupling model was 0.875 6, demonstrating the highest alignment with the distribution of existing hazard sites and accurately reflecting regional geological conditions. The findings could provide valuable references for future disaster prevention and mitigation in this area.

       

    /

    返回文章
    返回