Work

Enhanced groundwater potential mapping using a GIS based chaotic sparrow search algorithm optimized weighted broad learning system: A case study of the Guozhuang spring region, northern China

Year: 2025

Type: article

Abstract: Groundwater, as one of the most critical freshwater resources globally, is crucial for the sustainable utilization of water resources. However, the imbalance between scarce spring samples and abundant non-spring samples poses significant challenges for scientifically assessing groundwater potential. To address this issue, this paper proposes an improved weighted broad learning system (WBLS), integrated with an enhanced chaotic sparrow search algorithm (CSSA), aimed at tackling data imbalance and improving model prediction accuracy. This study focuses on the Guozhuang spring area in Shanxi Province, China. A dataset with a 1:10 ratio of positive to negative samples was constructed through GIS analysis and field surveys. This dataset was then divided into a training set (70%) and a testing set (30%). Influencing factors, including geology, human activities, and hydrology, were extracted from existing data. Through frequency ratio (FR) analysis, variable importance assessment, and multicollinearity evaluation, eleven key factors were ultimately identified. This study employed seven models: support vector machine (SVM), BLS, WBLS, SSA-BLS, CSSA-BLS, SSA-WBLS, and CSSA-WBLS to construct groundwater potential maps. The predictive performance of these models was evaluated by estimating the area under the receiver operating characteristic (ROC) curve (AUC) and several statistical metrics, including accuracy, sensitivity, specificity, balanced accuracy, and F1 score. The results showed that WBLS, SSA-WBLS, and CSSA-WBLS performed well in addressing data imbalance, with the CSSA-WBLS model achieving the best performance across all evaluation metrics. The GIS-generated groundwater potential map, derived from the predictive results, indicates that 18.78% of the Guozhuang spring area exhibits high groundwater potential. Overall, the CSSA-WBLS model demonstrated significant advantages in groundwater potential assessment. Therefore, under favorable conditions, this method holds considerable promise for application in various geographical environments.

Source: Journal of Hydrology Regional Studies

DOI: 10.1016/j.ejrh.2025.102708