Chen Meng, Zhang Shuangxi, Liu Shengbo, Li Mengkui, Zhang Tao, Wu Tengfei, Bu Xiangyu
School of Geodesy and Geomatics, Wuhan University, Wuhan, 430079, China.
Office of Research, City University of Wuhan, Wuhan, 430083, China.
Sci Rep. 2025 May 17;15(1):17159. doi: 10.1038/s41598-025-01837-y.
Rapid identification of groundwater sources is crucial for emergency water supplies. Yudu County (YDC) in Southern China serves as a case study due to its typical mountainous terrain and pressing groundwater demands. To address the limitations of conventional groundwater mapping methods in large-scale areas with sparse data, this study integrates remote sensing (RS), geographic information systems (GIS), and multi-criteria decision analysis (MCDA) techniques to delineate groundwater potential zones (GWPZs) in YDC. Following a series of correlation tests, seven assessment indicators were selected from various groundwater influencing factors, including two innovative ones: terrestrial water storage change (TWSC) and spring flow. The analytic hierarchy process (AHP) and fuzzy AHP (FAHP) models were employed to calculate factor weights, and GWPZ maps were generated using weighted overlay analysis in GIS. The model performance was validated using borewell data, receiver operating characteristic (ROC) curves, and yield prediction models. Additionally, four water enrichment types and their spatial distribution were identified by field investigations and yield prediction assessments. Results indicated a remarkable similarity between GWPZs delineated by AHP and FAHP, categorized into five classes: very high (13.92% for AHP and 14.33% for FAHP), high (26.29 and 27.55%), medium (29.33 and 28.14%), low (20.66 and 21.50%), and very low (9.80 and 8.48%). The area under the curve (AUC) for FAHP was 85.09%, slightly higher than the 84.41% of AHP, while the correlation coefficient (R of the prediction model improved from 0.747 to 0.817 with FAHP. These findings confirmed the reliability of combining GIS and MCDA methods to delineate GWPZs, with FAHP demonstrating an advantage over AHP. The proposed methodology and resulting mapping significantly enhance sustainable water resource management and development in YDC, offering a practical framework for rapid groundwater investigations in disaster response, as well as for long-term water security planning in similar mountainous environments.
快速识别地下水源对于应急供水至关重要。中国南方的于都县（YDC）因其典型的山区地形和紧迫的地下水需求而成为一个案例研究对象。为了解决传统地下水测绘方法在数据稀疏的大规模区域的局限性，本研究整合了遥感（RS）、地理信息系统（GIS）和多准则决策分析（MCDA）技术，以划定于都县的地下水潜力区（GWPZs）。经过一系列相关性测试，从各种地下水影响因素中选择了七个评估指标，其中包括两个创新指标：陆地水储量变化（TWSC）和泉水流量。采用层次分析法（AHP）和模糊层次分析法（FAHP）模型计算因子权重，并在GIS中使用加权叠加分析生成GWPZ地图。利用钻孔数据、接收器操作特征（ROC）曲线和产量预测模型对模型性能进行了验证。此外，通过现场调查和产量预测评估确定了四种富水类型及其空间分布。结果表明，AHP和FAHP划定的GWPZs之间具有显著相似性，分为五类：极高（AHP为13.92%，FAHP为14.33%）、高（26.29%和27.55%）、中（29.33%和28.14%）、低（20.66%和21.50%）和极低（9.80%和8.48%）。FAHP的曲线下面积（AUC）为85.09%，略高于AHP的84.41%，而预测模型的相关系数（R）从AHP的0.747提高到FAHP的0.817。这些发现证实了结合GIS和MCDA方法划定GWPZs的可靠性，FAHP显示出优于AHP的优势。所提出的方法和生成的地图显著加强了于都县的可持续水资源管理和开发，为灾害应对中的快速地下水调查以及类似山区环境中的长期水安全规划提供了一个实用框架。
