基于GIS与证据权模型的明光市地下水潜力分析

Analysis of groundwater potential in Mingguang City based on GIS and weights-of-evidence model

文章以安徽省明光市为研究区,选取高程、坡度、坡向、平面曲率、剖面曲率、地形湿度指数(topographic wetness index,TWI)、河网密度、公路缓冲区、断裂缓冲区、水系缓冲区和岩性特征为地下水潜力分析(groundwater potential analysis,GWPA)评价因子;基于现场调查数据,运用多重共线性分析方法优化评价因子,利用训练数据集构建证据权(weights-of-evidence,WoE)模型;在ArcGIS软件中通过空间叠加方法完成对研究区GWPA,将研究区内地下水潜力分为低、中、较高、高4类。结果表明:该方法具有较好的精度,方法简单易操作;研究区中、高潜力区占比分别为34.76%、22.98%,模型的接受者工作特征(receiver operating characteristic,ROC)曲线下面积值为0.7835;评价结果与验证数据集特征相符合,可为区域水资源管理提供参考。

This paper takes Mingguang City,Anhui Province as the research area,and selects altitude,slope,slope aspect,plane curvature,profile curvature,topographic wetness index(TWI),river network density,highway buffer zone,fault buffer zone,water system buffer zone,and lithological characteristics as evaluation factors for groundwater potential analysis(GWPA).Based on field survey data,multicollinearity analysis method is used to optimize evaluation factors,and training datasets are used to construct weights-of-evidence(WoE)model.The GWPA in the study area is completed by the spatial overlay method in ArcGIS software,and the groundwater potential in the study area is divided into four categories:low,medium,relatively high,and high.The results show that the method has good accuracy,and it is simple and easy to operate;the proportions of the medium and high potential areas in the study area are 34.76%and 22.98%,respectively,and the value of area under the receiver operating characteristic(ROC)curve of the model is 0.7835.The evaluation results are consistent with the characteristics of the validation dataset,which can provide a reference for regional water resources management.