基于OAT-GIS技术的地下水污染风险评估指标权重敏感性分析

Index Weight Sensitivity Analysis for Groundwater Pollution Risk Assessment Based on OAT–GIS

  • 摘要: 权重赋值准确性是地下水污染风险评估结论可靠性的重要基础,分析并量化指标权重敏感性是验证权重赋值准确性的难点。以我国37个危险废物填埋场为研究对象,基于构建的填埋场地下水污染风险排序指标体系,利用多准则决策分析模型(MCDA)中的层次分析(AHP)模块对各指标进行权重赋值;利用计算所得的权重最小变化量,基于OAT(one-factor-at-a-time)方法,应用MCDA的权重敏感性分析模块对14项指标权重进行依次改变,得出14项指标的权重敏感性大小;分析指标权重赋值过程中的不确定性,并基于地理信息系统(GIS)平台,开展可视化的空间数据统计与分析。结果表明:我国37个危险废物填埋场地下水污染风险评估的14项指标中,使用年限、渗滤液量和地形坡度3项指标权重改变后,填埋场地下水污染风险排序变化最大,即这3项指标权重敏感性最高;而包气带渗透性系数指标权重改变导致的风险排序变化最小,说明其权重敏感性最低。该方法可以有效地分析、验证指标权重赋值的准确性,识别权重赋值的不确定性,帮助相关决策者有针对性地进行风险管理。

     

    Abstract: The accuracy of the weight assignment is an important basis for the conclusion reliability of groundwater pollution risk assessment, and analyzing or quantifying the sensitivity of the index weight is the difficulty for verifying the accuracy of the weight assignment. Taking 37 hazardous waste landfills in China for case studies and based on the index system being built for groundwater pollution risk ranking, the weight of each index was assigned by using Analytic Hierarchy Process of multi-criteria decision analysis (MCDA) model. Using the calculated minimum changes of weight, the weights of 14 indices sequentially changed in the module of weight sensitivity analysis (WSA) by using MCDA model based on one-factor-at-a-time (OAT) method. It quantified the weight susceptibility strength of 14 indices and analyzed the uncertainty in the weight assignment process of the index. Finally, it launched the visualization of spatial data statistics and analysis based on GIS. The results showed that, after weight change of the three indices including useful life, leachate quantity and terrain slope, among the 14 indices used for the groundwater pollution risk assessment in 37 hazardous waste landfills, the risk ranking of groundwater pollution in landfill changed greatest, indicating that its weight sensitivity is highest. The weight change of vadose zone permeability coefficient leaded to a minimum change of the risk ranking, and it illustrated that the weight sensitivity of this index is minimum. The method can effectively analyze and validate the accuracy of index weight assignment, identify the uncertainty of weights assigned, and help relevant decision makers have specific risk management.

     

/

返回文章
返回