Abstract: Ancient stone terrace walls aimed at harvesting water runoff and facilitating crop production are widespread across the drylands of the Middle East and beyond. In addition to retaining the scarce water resource, the terrace walls also conserve soil and thicken its profile along ephemeral stream channels (wadis) by decreasing fluvial connectivity and mitigating erosional processes. In this study, we created hydrological models for three wadis with ancient stone terrace walls in the arid northern Negev of Israel, where the predominant geophyte species is Asphodelus ramosus L. A two-dimensional (2D) rain-on-grid (RoG) approach with a resolution of 2 m was used to simulate the rain events with return periods of 10, 20, 50, and 99 % (10-y, 5-y, 2-y, and yearly, respectively) based on the Intensity-Duration-Frequency rain curves for the region. To evaluate the effect of stone terrace walls on fluvial hydrology and geomorphology, the ground level was artificially elevated by 20 cm at the wall locations in a digital terrain model (DTM), using the built-in HEC-RAS 2D terrain modification tool. Our results showed that the terraced wadis have a high capacity to mitigate runoff loss, but a lesser capacity to delay the peak flow. Yet, for all rainstorm return periods, peak flow mitigation was positively related to the number of terrace walls along the stream channel. Field surveys in two of the studied wadis demonstrated that the A. ramosus clones were found in proximity to the stone terrace walls, presumably due to the greater soil–water content there. The results thus suggest that the terrace walls provide improved habitat conditions for these geophytes, supporting their growth and regulating their distribution along the wadi beds.

Abstract: The concentrations of U in natural waters are usually low, being typically less than 4 μg/L in river water, around 3.3 μg/L in open seawater, and usually less than 5 μg/L in groundwater. Higher concentrations can occur in both surface water and groundwater and the range spans some six orders of magnitude, with extremes in the mg/L range. However, such extremes in surface water are rare and linked to localized mineralization or evaporation in alkaline lakes. High concentrations in groundwater, substantially above the WHO provisional guideline value for U in drinking water of 30 μg/L, are associated most strongly with (i) granitic and felsic volcanic aquifers, (ii) continental sandstone aquifers especially in alluvial plains and (iii) areas of U mineralization. High-U groundwater provinces are more common in arid and semi-arid terrains where evaporation is an additional factor involved in concentrating U and other solutes. Examples of granitic and felsic volcanic terrains with documented high U concentrations include several parts of peninsular India, eastern USA, Canada, South Korea, southern Finland, Norway, Switzerland and Burundi. Examples of continental sandstone aquifers include the alluvial plains of the Indo-Gangetic Basin of India and Pakistan, the Central Valley, High Plains, Carson Desert, Española Basin and Edwards-Trinity aquifers of the USA, Datong Basin, China, parts of Iraq and the loess of the Chaco-Pampean Plain, Argentina. Many of these plains host eroded deposits of granitic and felsic volcanic precursors which likely act as primary sources of U. Numerous examples exist of groundwater impacted by U mineralization, often accompanied by mining, including locations in USA, Australia, Brazil, Canada, Portugal, China, Egypt and Germany. These may host high to extreme concentrations of U but are typically of localized extent. The overarching mechanisms of U mobilization in water are now well-established and depend broadly on redox conditions, pH and solute chemistry, which are shaped by the geological conditions outlined above. Uranium is recognized to be mobile in its oxic, U(VI) state, at neutral to alkaline pH (7–9) and is aided by the formation of stable U–CO3(±Ca, Mg) complexes. In such oxic and alkaline conditions, U commonly covaries with other similarly controlled anions and oxyanions such as F, As, V and Mo. Uranium is also mobile at acidic pH (2–4), principally as the uranyl cation UO22+. Mobility in U mineralized areas may therefore occur in neutral to alkaline conditions or in conditions with acid drainage, depending on the local occurrence and capacity for pH buffering by carbonate minerals. In groundwater, mobilization has also been observed in mildly (Mn-) reducing conditions. Uranium is immobile in more strongly (Fe-, SO4-) reducing conditions as it is reduced to U(IV) and is either precipitated as a crystalline or ‘non-crystalline’ form of UO2 or is sorbed to mineral surfaces. A more detailed understanding of U chemistry in the natural environment is challenging because of the large number of complexes formed, the strong binding to oxides and humic substances and their interactions, including ternary oxide-humic-U interactions. Improved quantification of these interactions will require updating of the commonly-used speciation software and databases to include the most recent developments in surface complexation models. Also, given their important role in maintaining low U concentrations in many natural waters, the nature and solubility of the amorphous or non-crystalline forms of UO2 that result from microbial reduction of U(VI) need improved quantification. Even where high-U groundwater exists, percentage exceedances of the WHO guideline value are variable and often small. More rigorous testing programmes to establish usable sources are therefore warranted in such vulnerable aquifers. As drinking-water regulation for U is a relatively recent introduction in many countries (e.g. the European Union), testing is not yet routine or established and data are still relatively limited. Acquisition of more data will establish whether analogous aquifers elsewhere in the world have similar patterns of aqueous U distribution. In the high-U groundwater regions that have been recognized so far, the general absence of evidence for clinical health symptoms is a positive finding and tempers the scale of public health concern, though it also highlights a need for continued investigation.

Abstract: Predicting groundwater levels is pivotal in curbing overexploitation and ensuring effective water resource governance. However, groundwater level prediction is intricate, driven by dynamic nonlinear factors. To comprehend the dynamic interaction among these drivers, leveraging machine learning models can provide valuable insights. The drastic increase in computational capabilities has catalysed a substantial surge in the utilisation of machine learning-based solutions for effective groundwater management. The performance of these models highly depends on the selection of hyperparameters. The optimisation of hyperparameters is a complex process that often requires application-specific expertise for a skillful prediction. To mitigate the challenge posed by hyperparameter tuning’s problem-specific nature, we present an innovative approach by introducing the automated machine learning (AutoML-GWL) framework. This framework is specifically designed for precise groundwater level mapping. It seamlessly integrates the selection of best machine learning model and adeptly fine-tunes its hyperparameters by using Bayesian optimisation. We used long time series (1997-2018) data of precipitation, temperature, evaporation, soil type, relative humidity, and lag of groundwater level as input features to train the AutoML-GWL model while considering the influence of Land Use Land Cover (LULC) as a contextual factor. Among these input features, the lag of groundwater level emerged as the most relevant input feature. Once the model is trained, it performs well over the unseen data with a strong correlation of coefficient (R = 0.90), low root mean square error (RMSE = 1.22), and minimal bias = 0.23. Further, we compared the performance of the proposed AutoML-GWL with sixteen benchmark algorithms comprising baseline and novel algorithms. The AutoML-GWL outperforms all the benchmark algorithms. Furthermore, the proposed algorithm ranked first in Friedman’s statistical test, confirming its reliability. Moreover, we conducted a spatial distribution and uncertainty analysis for the proposed algorithm. The outcomes of this analysis affirmed that the AutoML-GWL can effectively manage data with spatial variations and demonstrates remarkable stability when faced with small uncertainties in the input parameters. This study holds significant promise in revolutionising groundwater management practices by establishing an automated framework for simulating groundwater levels for sustainable water resource management.