Abstract: The crucial role which groundwater resource plays in our environment and the overall well-being of all living things can not be underestimated. Nonetheless, mismanagement of resources, over-exploitation, inadequate supply of surface water and pollution have led to severe drought and an overall drop in groundwater resources’ levels over the past decades. To address this, a groundwater flow model and several mathematical data-driven models have been developed for forecasting groundwater levels. However, there is a problem of unavailability and scarcity of the on-site input data needed by the data-driven models to forecast the groundwater level. Furthermore, as a result of the dynamics and stochastic characteristics of groundwater, there is a need for an appropriate, accurate and reliable forecasting model to solve these challenges. Over the years, the broad application of Machine Learning (ML) and Artificial Intelligence (AI) models are gaining attraction as an alternative solution for forecasting groundwater levels. Against this background, this article provides an overview of forecasting methods for predicting groundwater levels. Also, this article uses ML models such as Regressions Models, Deep Auto-Regressive models, and Nonlinear Autoregressive Neural Networks with External Input (NARX) to forecast groundwater levels using the groundwater region 10 at Karst belt in South Africa as a case study. This was done using Python Mx. Version 1.9.1., and MATLAB R2022a machine learning toolboxes. Moreover, the Coefficient of Determination (R2);, Root Mean Square Error (RMSE), Mutual Information gain, Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), Mean Absolute Error (MAE), and the Mean Absolute Scaled Error (MASE)) models were the forecasting statistical performance metrics used to assess the predictive performance of these models. The results obtained showed that NARX and Support Vector Machine (SVM) have higher performance metrics and accuracy compared to other models used in this study.

Abstract: In the southern San Joaquin Valley (SJV) of California, an agriculturally productive region that relies on groundwater for irrigation and domestic water supply, the infiltration of produced water from oil reservoirs is known to impact groundwater due to percolation from unlined disposal ponds. However, previously documented impacts almost exclusively focus on salinity, while contaminant loadings commonly associated with produced water (e.g., radionuclides) are poorly constrained. For example, the infiltration of bicarbonate-rich produced waters can react with sediment-bound uranium (U), leading to U mobilization and subsequent transport to nearby groundwater. Specifically, produced water infiltration poses a particular concern for SJV groundwater, as valley-fill sediments are well documented to be enriched in geogenic, reduced U. Here, we analyzed monitoring well data from two SJV produced water pond facilities to characterize U mobilization and subsequent groundwater contamination. Groundwater wells installed within 2 km of the facilities contained produced water and elevated levels of uranium. There are \textgreater400 produced water disposal pond facilities in the southern SJV. If our observations occur at even a fraction of these facilities, there is the potential for widespread U contamination in the groundwaters of one of the most productive agricultural regions in the world.

Abstract: Acid in situ leaching (AISL) is a subsurface mining approach suitable for low-grade ores which does not generate tailings, and has been adopted widely in uranium mining. However, this technique causes an extremely high concentration of contaminants at post-mining sites and in the surroundings soon after the mining ceases. As a potential AISL remediation strategy, natural attenuation has not been studied in detail. To address this problem, groundwater collected from 26 wells located within, adjacent, upgradient, and downgradient of a post-mining site were chosen to analyze the fate of U(VI), SO42−, δ34S, and δ238U, to reveal the main mechanisms governing the migration and attenuation of the dominant contaminants and the spatio-temporal evolutions of contaminants in the confined aquifer of the post-mining site. The δ238U values vary from −0.07 ‰ to 0.09 ‰ in the post-mining site and from −1.43 ‰ to 0.03 ‰ around the post-mining site. The δ34S values were found to vary from 3.3 ‰ to 6.2 ‰ in the post-mining site and from 6.0 ‰ to 11.0 ‰ around the post-mining site. Detailed analysis suggests that there are large differences between the range of isotopic composition variation and the range of pollutants concentration distribution, and the estimated Rayleigh isotope fractionation factor is 0.9994–0.9997 for uranium and 1.0032–1.0061 for sulfur. The isotope ratio of uranium and sulfur can be used to deduce the migration history of the contaminants and the irreversibility of the natural attenuation process in the anoxic confined aquifer. Combining the isotopic fractionation data for U and S with the concentrations of uranium and sulfate improved the accuracy of understanding of reducing conditions along the flow path. The study also indicated that as long as the geological conditions are favorable for redox reactions, natural attenuation could be used as a cost-effective remediation scheme.