Nitrate concentrations tracking from multi-aquifer groundwater vulnerability zones: Insight from machine learning and spatial mapping

Abstract

Nitrate contamination in groundwater is a significant environmental concern that poses risks to human health and ecosystems. Several goals and targets of Sustainable Development Goals (SDGs) are related to water quality, pollution, and sustainable management of water resources, which can encompass nitrate contamination. This study conducted real fieldwork of groundwater samples at several locations in Al-Hassa, Saudi Arabia. Subsequently, experimental based on chromatography (IC) and inductively coupled plasma mass spectrometry (ICP-MS) to analyze several groundwater hydro-geochemical elements. The study aimed to employ spatial mapping and advanced standalone optimization learning, including Elman Neural Network (ELNN), Gaussian Process Regression (GPR) models as nonparametric kernel-based probabilistic, and Random Forests (RFs) for tracking and modelling the nitrate (NO3) (mg/L) concentration. The outcomes were validated using several performance indicators and 2D-graphical methods. The resultant NO3 concentration in Al-Hassa was 77.9 mg/L, and the lowest was 9.8 mg/L. Despite marginal accuracy being obtained for most model combinations, GPR-C2 proved merit and reliable for modelling NO3 concentration with Wolffman Index (WI)=0.99 and PBAIS=−0.0001. The finding indicated that Al-Hassa regions are highly prone to NO3 pollution, further confirmed by spatial mapping. The outcomes provide insight into crucial information and decision-making for groundwater pollution risk at Al-Hassa, Saudi Arabia.