Predicting Redox Conditions in Groundwater at a Regional Scale

Occurrence and distribution of atrazine in groundwater from agricultural areas in China

There has been heavy application of chemical pesticides to farmland worldwide. As a state with a heavy emphasis on farming, China has extensively applied chemical pesticides to farmlands. Atrazine (ATZ), one of the most widely used herbicide in the world, has been used in large quantities in weed control in many crops, which poses a relatively large threat to groundwater utilization and agricultural safety. A nationwide sampling and analysis of ATZ in 8146 groundwater monitoring wells from 10 main grain producing areas were conducted during 2019. The results showed that the detection rate of ATZ in groundwater was 40.79 %, with detected concentrations up to 2.86 μg/L. ATZ detection rate was significantly higher in phreatic aquifers than confined aquifer. The concentration of ATZ decreased significantly with the depth increasing, the detected depth of ATZ in groundwater was varied from 5 to 1100 m, and 70.27 % of the detected ATZ was distributed in the depth of 0-51.8 m. The detection rate of ATZ in most north regions lower than south regions in China, and ATZ with high concentration was mainly occurred in southern regions. ATZ is more likely to migrate into the groundwater in the water resource rich areas. The concentration of ATZ decreased with TDS increasing, 74.39 % of the ATZ was detected in groundwater with TDS concentration <1 g/L. The hydrochemical type of groundwater in ATZ-enrichment environment were mainly Ca-HCO3 and Mixed type. According to classification and regression tree analysis, the high ATZ detection rate were closely related with groundwater depth (<51.8 m) and high organic matter content.

Seasonally varying contributions of contemporaneous and lagged sources of instream total nitrogen and phosphorus load across the Illinois River basin

Quantifying nutrient sources in streams, their temporal and spatial variability, and drivers of that variability can support effective water resources management. Yet a lack of data and modeling capabilities has previously prevented comprehensive quantification across both space and time. Here a dynamic SPARROW (Spatially Referenced Regressions on Watershed attributes) model that accounts for a lagged delivery of nutrients to streams was developed and applied to simulate seasonal and source-specific total nitrogen (TN) and total phosphorus (TP) loads in streams across the Illinois River basin (IRB). Dynamic load predictions from 2000 through 2020 revealed that a third of the TN and a quarter of the TP instream load originated from non-point sources that were lagged in their delivery from land-application to streams by more than a season. This lagged mass was the largest overall TN source-which was estimated as a lagged expression of previous seasonal non-point sources including fertilizer, manure, atmospheric deposition and fixation, and urban land use. Treated wastewater effluent was the largest TP source exported from the basin, contributing 39 % of the TP load and 15 % of the TN load, and dominated the load in the upper Illinois River near Chicago. Loads in the lower river during this period, conversely, were attributed primarily to a mix of agricultural sources and their lagged fractions from headwater tributaries. Instream processes removed 10 % of the TN load while only 4 % of the TP load was removed during instream transport. With appropriate datasets, the models could be extended to other basins or time periods and used to forecast future seasonal nutrient loads.

Short-Term Groundwater Level Fluctuations Drive Subsurface Redox Variability

As global change processes modify the extent and functions of terrestrial-aquatic interfaces, the variability of critical and dynamic transitional zones between wetlands and uplands increases. However, it is still unclear how fluctuating water levels at these dynamic boundaries alter groundwater biogeochemical cycling. Here, we used high-temporal resolution data along gradients from wetlands to uplands and during fluctuating water levels at freshwater coastal areas to capture spatiotemporal patterns of groundwater redox potential (Eh). We observed that topography influences groundwater Eh that is higher in uplands than in wetlands; however, the high variability within TAI zones challenged the establishment of distinct redox zonation. Declining water levels generally decreased Eh, but most locations exhibited significant Eh variability, which is associated with rare instances of short-term water level fluctuations, introducing oxygen. The Eh-oxygen relationship showed distinct hysteresis patterns, reflecting redox poising capacity at higher Eh, maintaining more oxidizing states longer than the dissolved oxygen presence. Surprisingly, we observed more frequent oxidizing states in transitional areas and wetlands than in uplands. We infer that occasional oxygen entering specific wetland-upland boundaries acts as critical biogeochemical control points. High-resolution data can capture such rare yet significant biogeochemical instances, supporting redox-informed models and advancing the predictability of climate change feedback.

Modeling groundwater redox conditions at national scale through integration of sediment color and water chemistry in a machine learning framework

Redox conditions play a crucial role in determining the fate of many contaminants in groundwater, impacting ecosystem services vital for both the aquatic environment and human water supply. Geospatial machine learning has previously successfully modelled large-scale redox conditions. This study is the first to consolidate the complementary information provided by sediment color and water chemistry to enhance our understanding of redox conditions in Denmark. In the first step, the depth to the first redox interface is modelled using sediment color from 27,042 boreholes. In the second step, the depth of the first redox interface is compared against water chemistry data at 22,198 wells to classify redox complexity. The absence of nitrate containing water below the first redox interface is referred to as continuous redox conditions. In contrast, discontinuous redox conditions are identified by the presence of nitrate below the first redox interface. Both models are built using 20 covariate maps, encompassing diverse hydrologically relevant information. The first redox interface is modelled with a mean error of 0.0 m and a root-mean-squared error of 8.0 m. The redox complexity model attains an accuracy of 69.8 %. Results indicate a mean depth to the first redox interface of 8.6 m and a standard deviation of 6.5 m. 60 % of Denmark is classified as discontinuous, indicating complex redox conditions, predominantly collocated in clay rich glacial landscapes. Both maps, i.e., first redox interface and redox complexity are largely driven by the water table and hydrogeology. The developed maps contribute to our understanding of subsurface redox processes, supporting national-scale land-use and water management.

Complex hydrology and variability of nitrogen sources in a karst watershed

Streams draining karst areas with rapid groundwater transit times may respond relatively quickly to nitrogen reduction strategies, but the complex hydrologic network of interconnected sinkholes and springs is challenging for determining the placement and effectiveness of management practices. This study aims to inform nitrogen reduction strategies in a representative agricultural karst setting of the Chesapeake Bay watershed (Fishing Creek watershed, Pennsylvania) with known elevated nitrate contamination and a previous documented groundwater residence time of less than a decade. During baseflow conditions, streamflow did not increase with drainage area. Headwaters and the main stem lost substantial flow to sinkholes until eventually discharging along large springs downstream. Seasonal hydrologic conditions shift the flow and nitrogen load spatially among losing and gaining stream sections. A compilation of nitrogen source inputs with the geochemistry and the pattern of enrichment of δ15N and δ18O suggest that the nitrogen in streams and springs during baseflow represents a mixture of manure, fertilizer, and wastewater sources with low potential for denitrification. The pH and calcite saturation index increased along generalized flow paths from headwaters to springs and indicate shorter groundwater residence times in baseflow during the spring versus summer. Given the substantial investment in management practices, fixed monitoring sites could incorporate synoptic water sampling to properly monitor long-term progress and help inform management actions in karst watersheds. Although karst watersheds have the potential to respond to nitrogen reduction strategies due to shorter groundwater residence times, high nitrogen inputs, effectiveness of conservation practices, and release of legacy nutrients within the karst cavities could confound progress of water quality goals.

Predicting Redox Conditions in Groundwater at a National Scale Using Random Forest Classification

Redox conditions in groundwater may markedly affect the fate and transport of nutrients, volatile organic compounds, and trace metals, with significant implications for human health. While many local assessments of redox conditions have been made, the spatial variability of redox reaction rates makes the determination of redox conditions at regional or national scales problematic. In this study, redox conditions in groundwater were predicted for the contiguous United States using random forest classification by relating measured water quality data from over 30,000 wells to natural and anthropogenic factors. The model correctly predicted the oxic/suboxic classification for 78 and 79% of the samples in the out-of-bag and hold-out data sets, respectively. Variables describing geology, hydrology, soil properties, and hydrologic position were among the most important factors affecting the likelihood of oxic conditions in groundwater. Important model variables tended to relate to aquifer recharge, groundwater travel time, or prevalence of electron donors, which are key drivers of redox conditions in groundwater. Partial dependence plots suggested that the likelihood of oxic conditions in groundwater decreased sharply as streams were approached and gradually as the depth below the water table increased. The probability of oxic groundwater increased as base flow index values increased, likely due to the prevalence of well-drained soils and geologic materials in high base flow index areas. The likelihood of oxic conditions increased as topographic wetness index (TWI) values decreased. High topographic wetness index values occur in areas with a propensity for standing water and overland flow, conditions that limit the delivery of dissolved oxygen to groundwater by recharge; higher TWI values also tend to occur in discharge areas, which may contain groundwater with long travel times. A second model was developed to predict the probability of elevated manganese (Mn) concentrations in groundwater (i.e., ≥50 μg/L). The Mn model relied on many of the same variables as the oxic/suboxic model and may be used to identify areas where Mn-reducing conditions occur and where there is an increased risk to domestic water supplies due to high Mn concentrations. Model predictions of redox conditions in groundwater produced in this study may help identify regions of the country with elevated groundwater vulnerability and stream vulnerability to groundwater-derived contaminants.

N-Map: High-resolution groundwater N-retention mapping and modelling by integration of geophysical, geological, geochemical, and hydrological data

A key aspect of protecting aquatic ecosystems from agricultural nitrogen (N) is to locate (i) farmlands where nitrate leaches from the bottom of the root zone and (ii) denitrifying zones in the aquifers where nitrate is removed before entering the surface water (N-retention). N-retention affects the choice of field mitigation measures to reduce delivered N to surface water. Farmland parcels associated with high N-retention gives the lowest impact of the targeted field measures and vice versa. In Denmark, a targeted N-regulation approach is currently implemented on small catchment scale (approx. 15 km²). Although this regulatory scale is much more detailed than what has been used previously, it is still so large that regulation for most individual fields will be either over- or under-regulated due to large spatial variation in the N-retention. The potential cost reduction for farmers is of up to 20-30% from detailed retention mapping at the field scale compared to the current small catchment scale. In this study, we present a mapping framework (N-Map) for differentiating farmland according to their N-retention, which can be used for improving the effectiveness of targeted N-regulation. The framework currently only includes N-retention in the groundwater. The framework benefits from the incorporation of innovative geophysics in hydrogeological and geochemical mapping and modelling. To capture and describe relevant uncertainties a large number of equally probable realizations are created through Multiple Point Statistical (MPS) methods. This allows relevant descriptions of uncertainties of parts of the model structure and includes other relevant uncertainty measures that affects the obtained N-retention. The output is data-driven high-resolution groundwater N-retention maps, to be used by the individual farmers to manage their cropping systems due to the given regulatory boundary conditions. The detailed mapping allows farmers to use this information in the farm planning in order to optimize the use of field measures to reduce delivered agricultural N to the surface water and thereby lower the costs of the field measures. From farmer interviews, however, it is clear that not all farms will have an economic gain from the detailed mapping as the mapping costs will exceed the potential economic gains for the farmers. The costs of N-Map is here estimated to 5-7 €/ha/year plus implementation costs at the farm. At the society level, the N-retention maps allow authorities to point out opportunities for a more targeted implementation of field measures to efficiently reduce the delivered N-load to surface waters.

Predictions of Arsenic in Domestic Well Water Sourced from Alluvial Aquifers of the Western Great Basin, USA

Chronic exposure to high levels of arsenic in drinking water can have wide-ranging health effects and is a global health concern. The domestic well population of the western Great Basin (WGB) is at increased risk of exposure to arsenic due to the hydrologic, geologic, and climatic setting of the region. A logistic regression (LR) model was developed to predict the probability of elevated arsenic (≥5 μg/L) in alluvial aquifers and assess the potential geologic hazard level posed to domestic well populations. Alluvial aquifers are susceptible to arsenic contamination, which is a concern because they are the primary source of water for domestic well users of the WGB. The probability of elevated arsenic at a domestic well is strongly influenced by tectonic and geothermal variables, including the total Quaternary fault length in the hydrographic basin and the distance between the sampled well and a geothermal system. The model had an overall accuracy of 81%, sensitivity of 92%, and specificity of 55%. Results show a >50% probability of elevated arsenic in untreated well water for approximately 49 thousand (64%) alluvial-aquifer domestic well users in northern Nevada, northeastern California, and western Utah.

Predicting regional fluoride concentrations at public and domestic supply depths in basin-fill aquifers of the western United States using a random forest model

A random forest regression (RFR) model was applied to over 12,000 wells with measured fluoride (F) concentrations in untreated groundwater to predict F concentrations at depths used for domestic and public supply in basin-fill aquifers of the western United States. The model relied on twenty-two regional-scale environmental and surficial predictor variables selected to represent factors known to control F concentrations in groundwater. The testing model fit R² and RMSE were 0.52 and 0.78 mg/L. Comparisons of measured to predicted proportions of four F-concentrations categories (<0.7 mg/L, 0.7-2 mg/L, >2 mg/L - 4 mg/L, and > 4 mg/L) indicate that the model performed well at making regional-scale predictions. Differences between measured and predicted proportions indicate underprediction of measured F at values by between 4 and 20 mg/L, representing less than 1% of the regional scale predicted values. These residuals most often map to geographic regions where local-scale processes including evaporative discharge in closed basins or intermittent streams concentrate fluoride in shallow groundwater. Despite this, the RFR model provides spatially continuous F predictions across the basin-fill aquifers where discrete samples are missing. Further, the predictions capture documented areas that exceed the F maximum contaminant level for drinking water of 4 mg/L and areas that are below the oral-health benchmark of 0.7 mg/L. These predictions can be used to estimate fluoride concentrations in unmonitored areas and to aid in identifying geographic areas that may require further investigation at localized scales.

Upscaling of Denitrification Rates from Point to Catchment Scales for Modeling of Nitrate Transport and Retention

The spatial and temporal variability of denitrification makes it challenging to integrate conceptual, process-based understandings of nitrate transport and retention into numerical modeling at the catchment scale, although it is critical for the realism and predictive power of the model. In this study, we propose a novel approach where the conceptual understandings of the spatial structure of denitrification zones and the corresponding representative denitrification rates are transformed into a form that can be integrated into a multi-point statistical simulation framework. This is done by constructing a denitrification training image (TI) coupled to a geophysically based TI of the hydrogeological structure. The field observations and laboratory analyses of denitrification rates and the chemistry of water and sediment revealed that the study catchment's subsurface can be characterized by three zones: (1) the oxic zone with no nitrate reduction; (2) the slow-denitrification zone (mean of ln-transformed rate = -1.19 ± 0.52 mg N L^-1 yr^-1); and (3) the high-denitrification zone (mean of ln-transformed rate = 3.86 ± 1.96 mg N L^-1 yr^-1). The underlying controls on the spatial distribution of these zones and the representativeness of denitrification rates were investigated. Then, a TI illustrating the subsurface structure of the denitrification zone was constructed by synthesizing the results of these geochemical interpretations and the hydrogeology TI.

Detecting anomalous methane in groundwater within hydrocarbon production areas across the United States

Numerous geochemical approaches have been proposed to ascertain if methane concentrations in groundwater, [CH₄], are anomalous, i.e., migrated from hydrocarbon production wells, rather than derived from natural sources. We propose a machine-learning model to consider alkalinity, Ca, Mg, Na, Ba, Fe, Mn, Cl, sulfate, TDS, specific conductance, pH, temperature, and turbidity holistically together. The model, an ensemble of sub-models targeting one parameter pair per sub-model, was trained with groundwater chemistry from Pennsylvania (n=19,086) and a set of 16 analyses from putatively contaminated groundwater. For cases where [CH₄] ≥ 10 mg/L, salinity- and redox-related parameters sometimes show that CH₄ may have moved into the aquifer recently and separately from natural brine migration, i.e., anomalous CH₄. We applied the model to validation and hold-out data for Pennsylvania (n=4,786) and groundwater data from three other gas-producing states: New York (n=203), Texas (n=688), and Colorado (n=10,258). The applications show that 1.4%, 1.3%, 0%, and 0.9% of tested samples in these four states, respectively, have high [CH₄] and are ≥50% likely to have been impacted by gas migrated from exploited reservoirs. If our approach is indeed successful in flagging anomalous CH₄, we conclude that: i) the frequency of anomalous CH₄ (# flagged water samples / total samples tested) in the Appalachian Basin is similar in areas where gas wells target unconventional as compared to conventional reservoirs, and ii) the frequency of anomalous CH₄ in Pennsylvania is higher than in Texas + Colorado. We cannot, however, exclude the possibility that differences among regions might be affected by differences in data volumes. Machine learning models will become increasingly useful in informing decision-making for shale gas development.

Predicting algal blooms: Are we overlooking groundwater?

Significant advances in understanding and predicting freshwater algal bloom dynamics have emerged in response to both increased occurrence and financial burden of nuisance and harmful blooms. Several factors have been highlighted as key controls of bloom occurrence, including nutrient dynamics, local hydrology, climatic perturbations, watershed geomorphology, biogeochemistry, food-web control, and algal competition. However, a major research gap continues to be the degree to which groundwater inputs modulate microbial biomass production and food-web dynamics at the terrestrial-aquatic interface. We present a synthesis of groundwater related algal bloom literature, upon which we derive a foundational hypothesis: long residence times cause groundwater to be geochemically and biologically distinct from surface water, allowing groundwater inputs to modulate algal bloom dynamics (growth, decline, toxicity) through its control over in-stream water chemistry. Distinct groundwater chemistry can support or prevent algal blooms, depending on specific local conditions. We highlight three mechanisms that influence the impact of groundwater discharge on algal growth: 1) redox state of the subsurface, 2) extent of water-rock interactions, and 3) stability of groundwater discharge. We underscore that in testing hypotheses related to groundwater control over algal blooms, it is critical to understand how changes in land use, water management, and climate will influence groundwater dynamics and, thus, algal bloom probabilities. Given this challenge, we argue that advances in both modeling and data integration, including genomics data and integrated process-based models that capture groundwater dynamics, are needed to illuminate mechanistic controls and improve predictions of algal blooms.

From Hydrometeorology to River Water Quality: Can a Deep Learning Model Predict Dissolved Oxygen at the Continental Scale?

Dissolved oxygen (DO) reflects river metabolic pulses and is an essential water quality measure. Our capabilities of forecasting DO however remain elusive. Water quality data, specifically DO data here, often have large gaps and sparse areal and temporal coverage. Earth surface and hydrometeorology data, on the other hand, have become largely available. Here we ask: can a Long Short-Term Memory (LSTM) model learn about river DO dynamics from sparse DO and intensive (daily) hydrometeorology data? We used CAMELS-chem, a new data set with DO concentrations from 236 minimally disturbed watersheds across the U.S. The model generally learns the theory of DO solubility and captures its decreasing trend with increasing water temperature. It exhibits the potential of predicting DO in "chemically ungauged basins", defined as basins without any measurements of DO and broadly water quality in general. The model however misses some DO peaks and troughs when in-stream biogeochemical processes become important. Surprisingly, the model does not perform better where more data are available. Instead, it performs better in basins with low variations of streamflow and DO, high runoff-ratio (>0.45), and winter precipitation peaks. Results here suggest that more data collections at DO peaks and troughs and in sparsely monitored areas are essential to overcome the issue of data scarcity, an outstanding challenge in the water quality community.

Factors Affecting Nitrate Concentrations in Stream Base Flow

Elevated nitrogen concentrations in streams and rivers in the Chesapeake Bay watershed have adversely affected the ecosystem health of the bay. Much of this nitrogen is derived as nitrate from groundwater that discharges to streams as base flow. In this study, boosted regression trees (BRTs) were used to relate nitrate concentrations in base flow (n = 156) to explanatory variables describing nitrogen sources, geology, and soil and catchment characteristics. From these relations, a BRT model was developed to predict base flow nitrate concentrations in streams throughout the Chesapeake Bay watershed. The highest base flow nitrate concentrations were associated with intensive agricultural land use, carbonate geology, and sparse riparian canopy, which suggested that reduced nitrogen inputs, particularly over carbonate terrane, are critical for limiting nitrate concentrations. The lowest nitrate concentrations in the BRT model were associated with extensive riparian canopy, high levels of organic carbon in soils, and suboxic conditions at shallow depths, which suggested that denitrification in the subsurface, particularly in the riparian zone, is limiting base flow nitrate concentrations. Nitrate transport from aquifers to streams can take decades to occur, resulting in decades-long lag times between the time when a land-use activity is implemented and when its effects are fully observed in streams. Predictive models of base flow nitrate concentrations in streams will help identify which portions of a watershed are likely to have large fractions of total stream nitrogen load derived from pathways with significant lag times.

The Shallow and Deep Hypothesis: Subsurface Vertical Chemical Contrasts Shape Nitrate Export Patterns from Different Land Uses

Eutrophication has threatened water resources worldwide, yet mechanistic understanding on controls of nutrient export remains elusive. This work tests the shallow and deep hypothesis: subsurface vertical chemical contrasts regulate nitrate export patterns under different land use conditions. We synthesized data from 228 watersheds and used reactive transport modeling (500 simulations) under broad land use, climate, and geology conditions. Data synthesis indicated that human perturbation has amplified chemical contrasts in shallow water (e.g., soil water) versus deep waters (e.g., groundwater), inducing primarily flushing patterns (concentrations increase with streamflow) in agriculture lands and dilution patterns (concentrations decrease with streamflow) in urban watersheds. Results revealed a quantitative relationship between export patterns and shallow-versus-deep concentration contrasts, underscoring the often-overlooked role of nutrient distribution over depth. Results challenge the commonly held perception that legacy stores in agricultural lands induce chemostasis where concentrations vary negligibly with streamflow. They suggest that nitrate concentrations from agricultural lands will escalate during large hydrological events, which can exacerbate nutrient export problems as flooding events intensify in the future climate.

Manganese (Mn) Concentrations and the Mn-Fe Relationship in Shallow Groundwater: Implications for Groundwater Monitoring

Manganese (Mn) concentrations in approximately 32,000 groundwater analyses from more than 4800 monitoring wells in northern Germany were evaluated. This region was considered well suited to study Mn in shallow groundwater in unconsolidated sediments. Spearman rank correlation was used to correlate between redox-sensitive parameters and the Mann–Kendall test for an evaluation of temporal trends. Manganese concentrations varied over two orders of magnitude and more than 40% of the wells had concentrations above 0.3 mg/L. Median Mn concentrations in the major hydrogeological units, the Geesten, tidal wetlands, and fluviatile lowlands were 0.12 mg/L, 0.46 mg/L, and 0.27 mg/L, respectively. Separating the data by land use, the median concentrations were 0.20 mg/L for arable land, 0.15 mg/L for forests, and 0.24 for grassland. Calculated background concentrations of Mn varied from <0.25 mg/L to 4.79 mg/L. A new parameter, ∆Mn-Fe, defined as the concentration difference between Mn and Fe in mg/L together with nitrate concentrations exceeding 50 mg/L was used to identify the fertilizer-borne input of nitrate. However, the factor controlling Mn occurrence seemingly was the depth of monitoring wells and the screen-length. Elevated concentrations of Mn and a high ∆Mn-Fe were generally found in shallow wells and wells with short screen-lengths.

Characterization and comparison of groundwater quality and redox conditions in the Arakawa Lowland and Musashino Upland, southern Kanto Plain of the Tokyo Metropolitan area, Japan

Groundwater is essential for the Earth biosphere but is often contaminated by harmful chemical compounds due to both anthropogenic and natural causes. A key factor controlling the fate of harmful chemicals in groundwater is the reduction/oxidation (redox) conditions. The formation factors for the groundwater redox conditions are insufficiently understood. In this study, long-term groundwater quality beneath one of the world megacities was monitored and evaluated. We measured and compared hydrogeochemical conditions including groundwater quality (35 chemical parameters) and redox conditions of five aquifers in the Arakawa Lowland and Musashino Upland, southern Kanto Plain of the Tokyo Metropolitan area, Japan. Monitoring results suggested the following: The main origin of groundwater is precipitation in both the Lowland and Upland areas. The three aquifers in the Arakawa Lowland are likely fully separated, with one unconfined and two confined aquifers under iron reducing and methanogenic conditions, respectively. Oppositely, in the Musashino Upland, the water masses in the two aquifers are likely partly connected, under aerobic conditions, and undergoing the same groundwater recharge and flow processes under similar hydrogeological conditions. The different groundwater redox conditions observed are likely caused by the very different groundwater residence times for the Arakawa Lowland and Musashino Upland.

Achieving unbiased predictions of national-scale groundwater redox conditions via data oversampling and statistical learning

An important policy consideration for integrated land and water management is to understand the spatial distribution of nitrate attenuation in the groundwater system, for which redox condition is the key indicator. This paper proposes a methodology to accommodate the computational demands of large datasets, and presents national-scale predictions of groundwater redox class for New Zealand. Our approach applies statistical learning methods to relate the redox class determined on groundwater samples to spatially varying attributes. The trained model uses these spatial variables to predict redox status in areas without sample data. We assembled the groundwater sample data from regional authority databases, and assigned each sample a redox class. A key achievement was to overcome the influence of sample selection bias on model training via oversampling. We removed additional bias imposed by imbalances in the predictor variables by applying a conditional inference random forest classifier. The unbiased trained model uses eight predictors, and achieves a high validation performance (accuracy 0.81, kappa 0.71), providing good confidence in model predictions. National maps are provided for redox class and probability at specified depths. Feature importance rankings indicate that reducing conditions are associated with poorly-drained soils, and to a lesser extent, high hydrological variability, low elevation, and low-permeability lithology. These conditions are common in New Zealand's coastal and lowland plains, where artificial drainage is required to make land suitable for production. The spatial extent of reduced groundwater increases with depth, suggesting a shallow influence of soil infiltration or mobile organic carbon, and a deeper influence of lithological electron donors. Our model provides unbiased predictions at a scale relevant for environmental policy development and legislation. Identifying where the ecosystem service provided by denitrification can be utilised will enable spatially targeted interventions that can achieve the desired environmental outcome in a more cost-effective manner than non-targeted interventions.

Drinking water quality in the glacial aquifer system, northern USA

Groundwater supplies 50% of drinking water worldwide, but compromised water quality from anthropogenic and geogenic contaminants can limit usage of groundwater as a drinking water source. Groundwater quality in the glacial aquifer system, USA (GLAC), is presented in the context of a hydrogeologic framework that divides the study area into 17 hydrogeologic terranes. Results are reported at aquifer-system scale and regional (terrane) scale. This paper presents a quantitative assessment of groundwater quality in the GLAC using data from numerous sources for samples collected 2005-2013, compared to health-based and aesthetic (non-health) benchmarks, and evaluated with areal and population metrics. Concentrations above a benchmark are considered high. Trace elements are widespread across the study area, with an estimated 5.7 million people relying on groundwater with high concentrations of one or more trace elements; manganese and arsenic are most often at high concentration. Nitrate is found at high concentration in 4.0% of the study area, serving about 740 thousand people. Organic compounds including pesticides and volatile organic compounds are high in 2.0% of the assessed study area, with about 870 thousand people relying on groundwater with high concentrations of an organic compound. High arsenic and manganese concentrations occur primarily in the terranes with thick, stratigraphically complex, fine-grained glacial sediment, coincident with groundwater under reducing conditions (indicated by iron concentrations >100 μg/L); high nitrate is uncommon in those same terranes. When nitrate is high in thick, fine-grained, complex terranes, though, it is much more commonly associated with groundwater under more oxidizing conditions. Common geogenic trace elements occur at high concentration due to characteristic geologic and geochemical conditions. Conversely, anthropogenic nitrate and organic compounds are introduced at or near the land surface. High concentrations of nitrate or organic compounds are generally limited to areas in proximity where people live and use the chemicals.

A Probabilistic Approach for Predicting Methane Occurrence in Groundwater

Aqueous geochemistry datasets from regional groundwater monitoring programs can be a major asset for environmental baseline assessment (EBA) in regions with development of natural gases from unconventional hydrocarbon resources. However, they usually do not include crucial parameters for EBA in areas of shale gas development such as methane concentrations. A logistic regression (LR) model was developed to predict the probability of methane occurrence in aquifers in Alberta (Canada). The model was calibrated and tested using geochemistry data including methane concentrations from two groundwater monitoring programs. The LR model correctly predicts methane occurrence in 89.8% (n = 234 samples) and 88.1% (n = 532 samples) of groundwater samples from each monitoring program. Methane concentrations strongly depend on the occurrence of electron donors such as sulfate and to a lesser extent on well depth and the total dissolved solids of groundwater. The model was then applied to a province-wide public health groundwater monitoring program (n = 52,849 samples) providing aqueous geochemistry data but no methane concentrations. This approach allowed the prediction of methane occurrence in regions where no groundwater gas data are available, thereby increasing the resolution of EBA in areas of shale gas development by using basic hydrochemical parameters measured in high-density groundwater monitoring programs.

The effects of denitrification parameterization and potential benefits of spatially targeted regulation for the reduction of N-discharges from agriculture

Diffuse nitrate leaching from agricultural areas is a major environmental problem in many parts of the world. Understanding where in a catchment nitrate is removed is key for designing effective land use management strategies that protect water quality, while minimizing the impact on economic development. In this study we assess the effects of spatially targeted nitrate leaching regulation in a basin with limited knowledge of the complexity of chemical heterogeneity. Three alternative nitrate reactivity spatial parameterizations were incorporated in a catchment-scale flow and transport model and used to evaluate the effectiveness of four possible spatially targeted regulation options. Our findings confirm that denitrification parameterization cannot be numerically determined based on model inversion alone. Detailed field based characterization using physical and geochemical methods should be considered and incorporated in the numerical inversion scheme. We also demonstrate that there are potential benefits of implementing spatially targeted regulation compared to spatially uniform regulation. Focusing regulation in areas where nitrate residence time is short, such as riparian zones or areas with low natural N-reduction, results in greater reduction of N-discharges through groundwater. Significantly improved efficiencies can be expected when delineation of management zones considers the chemical heterogeneity and groundwater flow paths. These improved efficiencies are achieved by adopting management rules that regulate land use in discharge sensitive areas, where leaching changes contribute the most to the catchment nitrate discharges. In our case study, regulation in discharge sensitive zones was twice as efficient compared to other management options.

Vertical stratification of redox conditions, denitrification and recharge in shallow groundwater on a volcanic hillslope containing relict organic matter

Natural denitrification in groundwater systems has been recognised as an ecosystem service that reduces the impact of agriculturally-derived nitrate inputs to surface waters. Identification of this ecosystem service within the landscape would permit spatially differentiated land management and legislation. However, spatial variation in groundwater redox conditions poses a significant challenge to such a concept. To gain understanding of the small-scale mosaic of biogeochemical and hydrological controls on denitrification, we established a well field consisting of 11 multilevel well (MLW) clusters on a hillslope containing relict organic matter buried by volcanic deposits 1.8 ka before present. Based on site-specific redox classification thresholds, vertical redox gradients and denitrification potentials were detected at 7 of the 11 sites. Palaeosols or woody debris, which had previously been identified in laboratory experiments as resident electron donors fuelling denitrification, were visually recognisable at 4 of the 7 MLW sites with vertical redox gradients. Moderately enhanced groundwater dissolved organic carbon (DOC) concentrations occurred where resident electron donors were evident. DOC concentrations were lower where anoxic and nitrate-depleted groundwater was found but with an absence of resident electron donors. In these instances, it was assumed that nitrate reduction had occurred somewhere upgradient of the sampled well screen along the lateral groundwater flow path, with the proximate electron donor (DOC) largely consumed in the process, since no evidence was found for denitrification being fuelled by inorganic electron donors. Due to high variability in the isotopic signature of nitrate in oxidised groundwater, the nitrate dual isotope method did not yield firm evidence for denitrification. However, realistic vertical patterns were obtained using the excess N₂ method. Tritium-based age dating revealed that oxic conditions were restricted to young groundwater (mean residence time ≤ 3 y), while anoxic conditions were observed across a wider age range (3-25 y).

Predicting Arsenic in Drinking Water Wells of the Central Valley, California

Probabilities of arsenic in groundwater at depths used for domestic and public supply in the Central Valley of California are predicted using weak-learner ensemble models (boosted regression trees, BRT) and more traditional linear models (logistic regression, LR). Both methods captured major processes that affect arsenic concentrations, such as the chemical evolution of groundwater, redox differences, and the influence of aquifer geochemistry. Inferred flow-path length was the most important variable but near-surface-aquifer geochemical data also were significant. A unique feature of this study was that previously predicted nitrate concentrations in three dimensions were themselves predictive of arsenic and indicated an important redox effect at >10 μg/L, indicating low arsenic where nitrate was high. Additionally, a variable representing three-dimensional aquifer texture from the Central Valley Hydrologic Model was an important predictor, indicating high arsenic associated with fine-grained aquifer sediment. BRT outperformed LR at the 5 μg/L threshold in all five predictive performance measures and at 10 μg/L in four out of five measures. BRT yielded higher prediction sensitivity (39%) than LR (18%) at the 10 μg/L threshold-a useful outcome because a major objective of the modeling was to improve our ability to predict high arsenic areas.