Physical models have long been employed for groundwater level (GWL) prediction. Recently, artificial intelligence (AI), particularly neural networks (NNs), has gained widespread use in forecasting GWL. Forecasting of GWL is essential to enable the analysis, quantifying, and management of groundwater. This systematic review investigates the application of NNs for GWL prediction, focusing on the architectures of the various NN models employed. The study utilizes the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) methodology to screen and synthesize relevant scientific articles. Various NN architectures, such as artificial neural networks (ANNs), feedforward neural networks (FFNNs), backpropagation neural networks (BPNNs), long short-term memory (LSTM), and hybrid models, were analyzed. The results from the systematic review indicate a growing preference for hybrid models, which effectively capture hidden relationships between GWL and environmental factors. The root mean square error (RMSE) emerges as the predominant performance metric, highlighting its significance in evaluating NNs. Results from the review also highlight the significance of comprehensive, long-term datasets covering a decade for robust trend analyses and accurate predictions. The findings contribute to a deeper understanding of new trends in groundwater research such as the application of neural networks for prediction problems in groundwater research. In conclusion, a hybrid metaheuristic algorithm produced more efficient results emphasizing their efficacy. In addition, lagged values were essential input for GWL prediction. The paper addressed both technical nuances and broader environmental implications.

The focus of groundwater research has evolved, expanded, and adapted to meet the water demands of society. In recent years, discernible trends have emerged in groundwater studies, particularly in the domains of exploration and exploitation. Groundwater exploration in West Africa has predominantly been driven by demand and ease of accessing water in its hosting environment. In light of the increasing dependence and the difficulty of groundwater accessibility in some environments, numerous methods have been employed to facilitate the rapid assessment of groundwater resources. This review explores into the diverse methods employed in groundwater exploration, placing particular emphasis on analyzing the evolution from geophysical methods to the application of machine learning in West Africa. The PRISMA technique was employed for data collection, screening, and verification of data eligibility. Researchers associated with Nigerian institutions contributed to 57% of all articles, followed by researchers from Ghanaian institutions. The predominant method employed in groundwater exploration, as evidenced by the highest number of published papers (18), was the electrical resistivity method followed by electromagnetic, and magnetic methodologies. Other techniques; such as induced polarization and gravity were employed in conjunction with other geophysical methods to enhance and/or corroborate the obtained results. Our research findings revealed that geophysical techniques played a crucial role in delineating the diverse geological environments relevant to groundwater accumulation and movement. In addition, Geographic Information Systems (GIS) and remote sensing have been integrated with geophysical methods to investigate the spatial distribution of groundwater resources. While these geophysical methods have demonstrated effectiveness, their interpretation relies on anomalies, which can be influenced by various factors. Incorrect interpretations may arise without thorough background studies on the geology of the area. Despite the fact that global trends are increasingly favoring the adoption of machine learning in groundwater exploration with great success in delineating potential groundwater reserves, our research reveals a notable scarcity of such endeavours in West Africa. Groundwater exploration in West Africa primarily revolves around the aforementioned geophysical methods and Geographic Information System (GIS). The review underscores the urgency for support, both financially and in terms of technological expertise, to promote research involving machine learning in West African countries. A key finding is the significant underrepresentation of machine learning studies by West African researchers in groundwater exploration, despite its demonstrated cost-effectiveness, time efficiency, and enhanced reliability.