Spatial Prediction of Groundwater Potentiality Mapping Using Machine Learning Algorithms

Abstract

Machine learning techniques offer powerful tools for the assessment and management of groundwater resources. Here, we evaluated the groundwater potential maps (GWPMs) in Md. Bazar Block of Birbhum District, India using four GIS-based machine-learning algorithms (MLA) such as predictive neural network (PNN), decision tree (DT), Naïve Bayes classifier (NBC), and random forest (RF). We used a database of 85 dug wells and one piezometer location identified using extensive field study, and employed 12 influencing factors (elevation, slope, drainage density (DD), topographical wetness index, geomorphology, lineament density, rainfall, geology, pond density, land use/land cover (LULC), geology, and soil texture) for evaluation through GIS. The 85 dug wells and 1 piezometer locations were sub-divided into two classes: 70:30 for training and model validation. The DT, RF, PNN, and NBC MLAs were implemented to analyse the relationship between the dug well locations and groundwater influencing factors to generate GWPMs. The results predict excellent groundwater potential areas (GPA) DT RF of 17.38%, 14.69%, 20.43%, and 13.97% of the study area, respectively. The prediction accuracy of each GWPM was determined using a receiver operating characteristic (ROC) curve. Using the 30% data sets (validation data), accuracies of 80.1%, 78.30%, 75.20%, and 69.2% were obtained for the PNN, RF, DT, and NBC models, respectively. The ROC values show that the four implemented models provide satisfactory and suitable results for GWP mapping in this region. In addition, the well-known mean decrease Gini (MDG) from the RF MLA was implemented to determine the relative importance of the variables for groundwater potentiality assessment. The MDG revealed that drainage density, lineament density, geomorphology, pond density, elevation, and stream junction frequency were the most useful determinants of GWPM. Our approach to delineate the GWPM can aid in the effective planning and management of groundwater resources in this region.

Drinking Water Resources Suitability Assessment Based on Pollution Index of Groundwater Using Improved Explainable Artificial Intelligence

“…Kulisz et al [24] tested the capacity of ANN methods to model the water quality index in groundwater and found satisfactory accuracy. Saha et al [25] evaluated groundwater potential maps (GWPMs) using machine learning algorithms (MLA) and achieved satisfactory results for groundwater potentiality assessment. They evaluated groundwater potential maps (GWPMs) using machine learning algorithms (MLA) and achieved satisfactory results for groundwater potentiality assessment [25].…”

“…Saha et al [25] evaluated groundwater potential maps (GWPMs) using machine learning algorithms (MLA) and achieved satisfactory results for groundwater potentiality assessment. They evaluated groundwater potential maps (GWPMs) using machine learning algorithms (MLA) and achieved satisfactory results for groundwater potentiality assessment [25]. Kumar et al [26] modelled groundwater drought indices using machine learning techniques, specifically ANN and random forest (RF), and found that the RF model showed superior performance [26].…”