Modeling groundwater nitrate concentrations in private wells in Iowa

A machine learning framework for spatio-temporal vulnerability mapping of groundwaters to nitrate in a data scarce region in Lenjanat Plain, Iran

The temporal aspect of groundwater vulnerability to contaminants such as nitrate is often overlooked, assuming vulnerability has a static nature. This study bridges this gap by employing machine learning with Detecting Breakpoints and Estimating Segments in Trend (DBEST) algorithm to reveal the underlying relationship between nitrate, water table, vegetation cover, and precipitation time series, that are related to agricultural activities and groundwater demand in a semi-arid region. The contamination probability of Lenjanat Plain has been mapped by comparing random forest (RF), support vector machine (SVM), and K-nearest-neighbors (KNN) models, fed with 32 input variables (dem-derived factors, physiography, distance and density maps, time series data). Also, imbalanced learning and feature selection techniques were investigated as supplementary methods, adding up to four scenarios. Results showed that the RF model, integrated with forward sequential feature selection (SFS) and SMOTE-Tomek resampling method, outperformed the other models (F1-score: 0.94, MCC: 0.83). The SFS techniques outperformed other feature selection methods in enhancing the accuracy of the models with the cost of computational expenses, and the cost-sensitive function proved more efficient in tackling imbalanced data issues than the other investigated methods. The DBEST method identified significant breakpoints within each time series dataset, revealing a clear association between agricultural practices along the Zayandehrood River and substantial nitrate contamination within the Lenjanat region. Additionally, the groundwater vulnerability maps created using the candid RF model and an ensemble of the best RF, SVM, and KNN models predicted mid to high levels of vulnerability in the central parts and the downhills in the southwest.

Nitrate exposure from drinking water and dietary sources among Iowa farmers using private wells

Nitrate levels are increasing in water resources across the United States and nitrate ingestion from drinking water has been associated with adverse health risks in epidemiologic studies at levels below the maximum contaminant level (MCL). In contrast, dietary nitrate ingestion has generally been associated with beneficial health effects. Few studies have characterized the contribution of both drinking water and dietary sources to nitrate exposure. The Agricultural Health Study is a prospective cohort of farmers and their spouses in Iowa and North Carolina. In 2018-2019, we assessed nitrate exposure for 47 farmers who used private wells for their drinking water and lived in 8 eastern Iowa counties where groundwater is vulnerable to nitrate contamination. Drinking water and dietary intakes were estimated using the National Cancer Institute Automated Self-Administered 24-Hour Dietary Assessment tool. We measured nitrate in tap water and estimated dietary nitrate from a database of food concentrations. Urinary nitrate was measured in first morning void samples in 2018-19 and in archived samples from 2010 to 2017 (minimum time between samples: 2 years; median: 7 years). We used linear regression to evaluate urinary nitrate concentrations in relation to total nitrate, and drinking water and dietary intakes separately. Overall, dietary nitrate contributed the most to total intake (median: 97 %; interquartile range [IQR]: 57-99 %). Among 15 participants (32 %) whose drinking water nitrate concentrations were at/above the U.S. Environmental Protection Agency MCL (10 mg/L NO3-N), median intake from water was 44 % (IQR: 26-72 %). Total nitrate intake was the strongest predictor of urinary nitrate concentrations (R2 = 0.53). Drinking water explained a similar proportion of the variation in nitrate excretion (R2 = 0.52) as diet (R2 = 0.47). Our findings demonstrate the importance of both dietary and drinking water intakes as determinants of nitrate excretion.

Mapping specific groundwater nitrate concentrations from spatial data using machine learning: A case study of chongqing, China

Groundwater resources is not only important essential water resources but also imperative connectors within the intricate framework of the ecological environment. High nitrate concentrations in groundwater can exerting adverse impacts on human health. It is imperative to accurately delineate the distribution characteristics of groundwater nitrate concentrations. Four different machine learning models (Gradient Boosting Regression (GB), Random Forest Regression (RF), Extreme Gradient Boosting Regression (XG) and Adaptive Boosting Regression (AD)) which combine spatial environmental data and different radius contributing area was developed to predict the distribution of nitrate concentration in groundwater. The models use 595 groundwater samples and included topography, remote sensing, hydrogeological and hydrological, climate, nitrate input, and socio-economic predictor. Gradient Boosting Regression model outperforms the other models (R2 = 0.627, MAE = 0.529, RMSE = 0.705, PICP = 0.924 for test dataset) under 500 m radius contributing area. A high-resolution (1 km) groundwater nitrate concentration distribution map reveal in the majority of the study area, groundwater nitrate concentrations are below 1 mg/L and high nitrate concentration (>10 mg/L) proportion in southeast, northeast and central main urban area karst valley regions is 1.89%, 0.91%, and 0.38% respectively. In study area, hydrogeological conditions, soil parameters, nitrogen input factors, and percentage of arable land are among the most influential explanatory factors. This work, serving as the inaugural application of utilizing effective spatial methods for predicting groundwater nitrate concentrations in Chongqing city, furnish decision-making support for the prevention and control of groundwater pollution, particularly in areas primarily dependent on groundwater for water supply and holds profound significance as a milestone achievement.

Ingested nitrate and nitrite and end-stage renal disease in licensed pesticide applicators and spouses in the Agricultural Health Study

BACKGROUND

Nitrate and nitrite ingestion has been linked to kidney cancer, possibly via the endogenous formation of carcinogenic N-nitroso compounds. These exposures might also contribute to end-stage renal disease (ESRD).

OBJECTIVES

We investigated associations of drinking water nitrate and dietary nitrate and nitrite intakes (total and by food type) with incident ESRD in the Agricultural Health Study. We also explored modifying effects of vitamin C and heme iron intake, which may affect endogenous nitrosation.

METHODS

We performed complete case analyses among private pesticide applicators and their spouses. We obtained water nitrate estimates for participants whose primary drinking water source at enrollment (1993-1997) was public water supplies (PWS) or private wells (N = 59,632). Average nitrate concentrations were computed from historical data for PWS users and predicted from random forest models for private well users. Analysis of dietary nitrate and nitrite was restricted to the 30,177 participants who completed the NCI Dietary History Questionnaire during follow-up (1999-2003). Incident ESRD through 2018 was ascertained through linkage with the U.S. Renal Data System. We estimated adjusted hazard ratios (HRs) and 95%CI for associations of tertiles (T) of exposure with ESRD overall and explored effects in strata of vitamin C and heme iron intake.

RESULTS

We identified 469 incident ESRD cases (206 for dietary analysis). Water nitrate and total dietary nitrate/nitrite were not associated with ESRD, but increased ESRD was associated with nitrate and nitrite from processed meats. We found apparent associations between nitrite and ESRD only among participants with vitamin C

SIGNIFICANCE

ESRD incidence was associated with dietary nitrate/nitrite from processed meat among all study participants and with total dietary nitrite among participants with lower vitamin C or higher heme iron intake.

IMPACT STATEMENT

There are few well-established environmental risk factors for end-stage renal disease (ESRD), a worldwide public health challenge. Ingestion of nitrate and nitrite, which may lead to endogenous formation of carcinogenic N-nitroso compounds, has been linked to some cancers and chronic diseases. We investigated these exposures in relation to ESRD in an agricultural cohort. ESRD incidence was associated with dietary nitrate/nitrite from processed meat and with total dietary nitrite among subgroups with lower vitamin C or higher heme iron intake. This study provides preliminary evidence that points to dietary nitrite and possibly dietary nitrate intake as a potential contributor to ESRD.

Collapse

Key Words

• Chronic kidney disease
• End-stage renal disease
• Nitrate
• Nitrite
• Processed meat
• Water pollution

Collapse

MESH Headings

• Humans
• Nitrates/analysis
• Kidney Failure, Chronic/epidemiology
• Kidney Failure, Chronic/chemically induced
• Male
• Nitrites/analysis
• Nitrites/adverse effects
• Female
• Middle Aged
• Spouses/statistics & numerical data
• Adult
• Pesticides
• Occupational Exposure/adverse effects
• Occupational Exposure/analysis
• Aged
• Drinking Water/chemistry
• Drinking Water/analysis
• Incidence
• Agriculture
• Diet/statistics & numerical data
• Farmers/statistics & numerical data
• Ascorbic Acid

Collapse

Grants Collapse

Affiliation(s)

Dazhe Chen Epidemiology Branch, National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC, USA
Christine G Parks Epidemiology Branch, National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC, USA
Laura E Beane Freeman Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Rockville, MD, USA
Jonathan N Hofmann Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Rockville, MD, USA
Rashmi Sinha Metabolic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Rockville, MD, USA
Jessica M Madrigal Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Rockville, MD, USA
Mary H Ward Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Rockville, MD, USA
Dale P Sandler Epidemiology Branch, National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC, USA.

Collapse

US drinking water quality: exposure risk profiles for seven legacy and emerging contaminants

BACKGROUND

Advances in drinking water infrastructure and treatment throughout the 20th and early 21st century dramatically improved water reliability and quality in the United States (US) and other parts of the world. However, numerous chemical contaminants from a range of anthropogenic and natural sources continue to pose chronic health concerns, even in countries with established drinking water regulations, such as the US.

OBJECTIVE/METHODS

In this review, we summarize exposure risk profiles and health effects for seven legacy and emerging drinking water contaminants or contaminant groups: arsenic, disinfection by-products, fracking-related substances, lead, nitrate, per- and polyfluorinated alkyl substances (PFAS) and uranium. We begin with an overview of US public water systems, and US and global drinking water regulation. We end with a summary of cross-cutting challenges that burden US drinking water systems: aging and deteriorated water infrastructure, vulnerabilities for children in school and childcare facilities, climate change, disparities in access to safe and reliable drinking water, uneven enforcement of drinking water standards, inadequate health assessments, large numbers of chemicals within a class, a preponderance of small water systems, and issues facing US Indigenous communities.

RESULTS

Research and data on US drinking water contamination show that exposure profiles, health risks, and water quality reliability issues vary widely across populations, geographically and by contaminant. Factors include water source, local and regional features, aging water infrastructure, industrial or commercial activities, and social determinants. Understanding the risk profiles of different drinking water contaminants is necessary for anticipating local and general problems, ascertaining the state of drinking water resources, and developing mitigation strategies.

IMPACT STATEMENT

Drinking water contamination is widespread, even in the US. Exposure risk profiles vary by contaminant. Understanding the risk profiles of different drinking water contaminants is necessary for anticipating local and general public health problems, ascertaining the state of drinking water resources, and developing mitigation strategies.

Improving groundwater nitrate concentration prediction using local ensemble of machine learning models

Groundwater is one of the most important water resources around the world, which is increasingly exposed to contamination. As nitrate is a common pollutant of groundwater and has negative effects on human health, predicting its concentration is of particular importance. Ensemble machine learning (ML) algorithms have been widely employed for nitrate concentration prediction in groundwater. However, existing ensemble models often overlook spatial variation by combining ML models with conventional methods like averaging. The objective of this study is to enhance the spatial accuracy of groundwater nitrate concentration prediction by integrating the outputs of ML models using a local approach that accounts for spatial variation. Initially, three widely used ML models including random forest regression (RFR), k-nearest neighbor (KNN), and support vector regression (SVR) were employed to predict groundwater nitrate concentration of Qom aquifer in Iran. Subsequently, the output of these models were integrated using geographically weighted regression (GWR) as a local model. The findings demonstrated that the ensemble of ML models using GWR resulted in the highest performance (R2 = 0.75 and RMSE = 9.38 mg/l) compared to an ensemble model using averaging (R2 = 0.68 and RMSE = 10.56 mg/l), as well as individual models such as RFR (R2 = 0.70 and RMSE = 10.16 mg/l), SVR (R2 = 0.59 and RMSE = 11.95 mg/l), and KNN (R2 = 0.57 and RMSE = 12.19 mg/l). The resulting prediction map revealed that groundwater nitrate contamination is predominantly concentrated in urban areas located in the northwestern regions of the study area. The insights gained from this study have practical implications for managers, assisting them in preventing nitrate pollution in groundwater and formulating strategies to improve water quality.

Quantifying the spatiotemporal variability of nitrate in irrigation water across the Wisconsin Central Sands

The Wisconsin Central Sands is home to large scale vegetable production on sandy soils and managed with frequent irrigation, fertigation, and widespread nitrogen fertilizer application, all of which make the region highly susceptible to nitrate loss to groundwater. While the groundwater is used as the primary source of drinking water for many communities and rural residences across the region, it is also used for irrigation. Considering the high levels of nitrate found in the groundwater, it has been proposed that growers more accurately account for the nitrate in their irrigation water as part of nitrogen management plans. Our objectives were to 1) determine the magnitude of nitrate in irrigation water, 2) quantify the spatiotemporal variability of nitrate, and 3) determine key predictors of nitrate concentration in the region. We sampled irrigation water from 38 fields across six farms from 2018 to 2020. Across the 3 years of our study, nitrate concentration varied more across space than time. On average, our samples were tested at 19.0 mg L-1 nitrate-nitrogen, or nearly two times the U.S. Environmental Protection Agency (EPA) threshold for safe drinking water, equivalent to 48.1 kg ha-1 of applied nitrate-nitrogen with 25.4 cm (or 10 in.) of irrigation. To better understand the spatiotemporal variability in nitrate levels, week of sampling, year, well depth, well casing, and nitrogen application rate were analyzed for their role as predictor variables. Based on our linear mixed effects model, nitrogen application rate was the greatest predictor of the nitrate concentration of irrigation water (p < 0.05).

Geostatistical estimates of groundwater nitrate-nitrogen concentrations with spatial auxiliary information on DRASTIC-LU-based aquifer contamination vulnerability

Groundwater nitrate-nitrogen contamination typically involves several natural and anthropogenic factors, including those related to hydrology, hydrogeology, topography, and land use (LU). DRASTIC-LU-based aquifer contamination vulnerability could be used to characterize the pollution potentials of groundwater nitrate-nitrogen and to determine groundwater protection zones. This study used regression kriging (RK) with environmental auxiliary information on DRASTIC-LU-based aquifer contamination vulnerability to investigate groundwater nitrate-nitrogen pollution in the Pingtung Plain of Taiwan. First, the relationship between groundwater nitrate-nitrogen pollution and assessments of aquifer contamination vulnerability was determined using stepwise multivariate linear regression (MLR). Subsequently, the residuals between the nitrate-nitrogen observations and MLR predictions were estimated by kriging techniques. Finally, the groundwater nitrate-nitrogen distributions were spatially analyzed using RK, ordinary kriging (OK), and MLR. The findings indicated that the land used for orchards and the medium- and coarse-sand fractions of vadose zones were associated with groundwater nitrate-nitrogen concentrations. The fertilizer used for orchards was identified as the primary source of groundwater nitrate-nitrogen pollution. The RK estimates could be used to analyze the characteristics of the pollution source for land used for orchards and exhibited high spatial variability and accuracy after residual correction. Moreover, RK had an excellent estimate ability for extreme data compared to MLR and OK. Correctly determining groundwater nitrate-nitrogen distributions using RK was useful for administering environmental resources and preventing public health hazards.

A deep learning-based novel approach to generate continuous daily stream nitrate concentration for nitrate data-sparse watersheds

High-frequency stream nitrate concentration provides critical insights into nutrient dynamics and can help to improve the effectiveness of management decisions to maintain a sustainable ecosystem. However, nitrate monitoring is conventionally conducted through lab analysis using in situ water samples and is typically at coarse temporal resolution. In the last decade, many agencies started collecting high-frequency (5-60 min intervals) nitrate data using optical sensors. The hypothesis of the study is that the data-driven models can learn the trend and temporal variability in nitrate concentration from high-frequency sensor-based nitrate data in the region and generate continuous nitrate data for unavailable data periods and data-limited locations. A Long Short-Term Memory (LSTM) model-based framework was developed to estimate continuous daily stream nitrate for dozens of gauge locations in Iowa, USA. The promising results supported the hypothesis; the LSTM model demonstrated median test-period Nash-Sutcliffe efficiency (NSE) = 0.75 and RMSE = 1.53 mg/L for estimating continuous daily nitrate concentration in 42 sites, which are unprecedented performance levels. Twenty-one sites (50 % of all sites) and thirty-four sites (76 % of all sites) demonstrated NSE > 0.75 and 0.50, respectively. The average nitrate concentration of neighboring sites was identified as a crucial determinant of continuous daily nitrate concentration. Seasonal model performance evaluation showed that the model performed effectively in the summer and fall seasons. About 26 sites showed correlations >0.60 between estimated nitrate concentration and discharge. The concentration-discharge (c-Q) relationship analysis showed that the study watersheds had four dominant nitrate transport patterns from landscapes to streams with increasing discharge, including the flushing pattern being the most dominant one. Stream nitrate estimation impedes due to data inadequacy. The modeling framework can be used to generate temporally continuous nitrate at nitrate data-limited regions with a nearby sensor-based nitrate gauge. Watershed planners and policymakers could utilize the continuous nitrate data to gain more information on the regional nitrate status and design conservation practices accordingly.

Application of machine learning in groundwater quality modeling - A comprehensive review

Groundwater is a crucial resource across agricultural, civil, and industrial sectors. The prediction of groundwater pollution due to various chemical components is vital for planning, policymaking, and management of groundwater resources. In the last two decades, the application of machine learning (ML) techniques for groundwater quality (GWQ) modeling has grown exponentially. This review assesses all supervised, semi-supervised, unsupervised, and ensemble ML models implemented to predict any groundwater quality parameter, making this the most extensive modern review on this topic. Neural networks are the most used ML model in GWQ modeling. Their usage has declined in recent years, giving rise to more accurate or advanced techniques such as deep learning or unsupervised algorithms. Iran and the United States lead the world in areas modeled, with a wealth of historical data available. Nitrate has been modeled most exhaustively, targeted by nearly half of all studies. Advancements in future work will be made with further implementation of deep learning and explainable artificial intelligence or other cutting-edge techniques, application of these techniques for sparsely studied variables, the modeling of new or unique study areas, and the implementation of ML techniques for groundwater quality management.

Air temperature spikes increase bacteria presence in drinking water wells downstream of hog lagoons

>44 million United States residents depend on private drinking water wells that are federally unregulated. Maintaining a clean groundwater supply for populations without access to public water systems is essential to supporting public health and falls to state regulators and private well owners. Yet, monitoring practices do not reflect the fact that groundwater pollution risk varies seasonally and with proximity to nearby surface-contaminated sites. Examination of nearly 50,000 well water samples across North Carolina, ranked second nationally in domestic well dependence and swine production, from 2013 to 2018 reveals a uniform sampling schedule but a variable risk of bacterial contamination within each calendar year. We document a threshold of 32.2 °C (90 °F) where total coliform bacteria and Escherichia coli (E. coli) detection in private well water spikes near swine lagoons but is absent from "upstream" wells and otherwise unexplained by a variety of other known contamination sites. Closing the gap between perceived and actual risks of drinking water contamination has potential to improve public health. State regulations and federal guidelines should consider coordinating domestic well sampling with seasonally and spatially fluctuating risks of groundwater contamination. Findings from this study are generalizable, having implications for other parts of the world with water sources that have the potential to get contaminated by nearby surface sources of human and animal waste, such as manure applications and leaching septic systems.

Prevalence of neonicotinoid insecticides in paired private-well tap water and human urine samples in a region of intense agriculture overlying vulnerable aquifers in eastern Iowa

A pilot study among farming households in eastern Iowa was conducted to assess human exposure to neonicotinoids (NEOs). The study was in a region with intense crop and livestock production and where groundwater is vulnerable to surface-applied contaminants. In addition to paired outdoor (hydrant) water and indoor (tap) water samples from private wells, urine samples were collected from 47 adult male pesticide applicators along with the completions of dietary and occupational surveys. Estimated Daily Intake (EDI) were then calculated to examine exposures for different aged family members. NEOs were detected in 53% of outdoor and 55% of indoor samples, with two or more NEOs in 13% of samples. Clothianidin was the most frequently detected NEO in water samples. Human exposure was ubiquitous in urine samples. A median of 10 different NEOs and/or metabolites were detected in urine, with clothianidin, nitenpyram, thiamethoxam, 6-chloronicotinic acid, and thiacloprid amide detected in every urine samples analyzed. Dinotefuran, imidaclothiz, acetamiprid-N-desmethyl, and N-desmethyl thiamethoxam were found in ≥70% of urine samples. Observed water intake for study participants and EDIs were below the chronic reference doses (CRfD) and acceptable daily intake (ADI) standards for all NEOs indicating minimal risk from ingestion of tap water. The study results indicate that while the consumption of private well tap water provides a human exposure pathway, the companion urine results provide evidence that diet and/or other exposure pathways (e.g., occupational, house dust) may contribute to exposure more than water contamination. Further biomonitoring research is needed to better understand the scale of human exposure from different sources.

The Utility of Machine Learning Models for Predicting Chemical Contaminants in Drinking Water: Promise, Challenges, and Opportunities

PURPOSE OF REVIEW

This review aims to better understand the utility of machine learning algorithms for predicting spatial patterns of contaminants in the United States (U.S.) drinking water.

RECENT FINDINGS

We found 27 U.S. drinking water studies in the past ten years that used machine learning algorithms to predict water quality. Most studies (42%) developed random forest classification models for groundwater. Continuous models show low predictive power, suggesting that larger datasets and additional predictors are needed. Categorical/classification models for arsenic and nitrate that predict exceedances of pollution thresholds are most common in the literature because of good national scale data coverage and priority as environmental health concerns. Most groundwater data used to develop models were obtained from the United States Geological Survey (USGS) National Water Information System (NWIS). Predictors were similar across contaminants but challenges are posed by the lack of a standard methodology for imputation, pre-processing, and differing availability of data across regions. We reviewed 27 articles that focused on seven drinking water contaminants. Good performance metrics were reported for binary models that classified chemical concentrations above a threshold value by finding significant predictors. Classification models are especially useful for assisting in the design of sampling efforts by identifying high-risk areas. Only a few studies have developed continuous models and obtaining good predictive performance for such models is still challenging. Improving continuous models is important for potential future use in epidemiological studies to supplement data gaps in exposure assessments for drinking water contaminants. While significant progress has been made over the past decade, methodological advances are still needed for selecting appropriate model performance metrics and accounting for spatial autocorrelations in data. Finally, improved infrastructure for code and data sharing would spearhead more rapid advances in machine-learning models for drinking water quality.

Gis-Based Assessment of Risk for Drinking Water Contamination to Children’s Health in Rural Settlements

Nitrates and heavy metals are two of the most significant groundwater pollutants that have an impact on people’s health all around the world. In order to assess their risk to children’s health, this study aims to determine the total iron, manganese, and nitrate content in drinking water sources of non-centralized water supply of the educational institutions in the rural settlements of Zhytomyr region. A total of 114 water samples from wells and bores used by the educational institutions to supply domestic and drinking water to rural residential areas were collected. The Measurement Laboratory of Polissia National University conducted analytical studies. The danger to children’s health was evaluated using the hazard quotient and following the methodology recommended by the US Environmental Protection Agency. The software ArcGIS Pro was used to identify risk zones.

The average total iron content in the drinking water of the Berdichev, Zhytomyr, and Novohrad-Volinsky districts was 1.5–2.8 times higher. In all regions, the average manganese concentration did not go above the allowable level. On average, the nitrate content was also below the threshold, but in 22.6–42.9% of the samples, it exceeded the allowable level. Children aged 6–12 years were at the highest risk, and children living in Berdichev district had the highest hazard quotient at 1.972. The fact that nitrates accounted for 67–84% of the total risk indicates the danger associated with the intake of nitrates, even in amounts below the allowable concentration.

Susceptibility Assessment of Groundwater Nitrate Contamination Using an Ensemble Machine Learning Approach

Groundwater pollution susceptibility mapping using parsimonious approaches with limited data is of utmost importance for water resource and health planning, especially in data-scarce regions. Current research assesses groundwater nitrate susceptibility by considering the various combination of explanatory variables. In this study, the novel machine learning models of weighted subspace random forest (WSRF) and generalized additive model using LOESS (GAMLOESS) are applied, and the results are compared with well-known machine learning models of K-nearest neighbors (KKNN) and random forest (RF). The optimum combination of inputs for groundwater nitrate susceptibility mapping is identified using the k-fold cross-validation methodology. Results indicated that the combination of variables of precipitation, groundwater level, and lithology had the best performance among the 16 combinations. Modeling performance using the optimum combination demonstrated that the new ensemble approach, the WSRF model, had superior performance according to the evaluation metrics of accuracy (0.87), kappa (0.73), precision (0.92), false alarm ratio (0.08), and critical success index (0.75). The susceptibility assessment results of this paper can be a useful tool in developing strategies for the prevention and protection of groundwater pollution.

Sources of ammonium enriched in groundwater in the central Yangtze River Basin: Anthropogenic or geogenic?

The occurrence of excessive ammonium in groundwater threatens human and aquatic ecosystem health across many places worldwide. As the fate of ammonium in groundwater systems is often affected by a complex mixture of transport and biogeochemical transformation processes, identifying the sources of groundwater ammonium is an important prerequisite for planning effective mitigation strategies. Elevated ammonium was found in both a shallow and an underlying deep groundwater system in an alluvial aquifer system beneath an agricultural area in the central Yangtze River Basin, China. In this study we develop and apply a novel, indirect approach, which couples the random forest classification (RFC) of machine learning method and fluorescence excitation-emission matrices with parallel factor analysis (EEM-PARAFAC), to distinguish multiple sources of ammonium in a multi-layer aquifer. EEM-PARAFAC was applied to provide insights into potential ammonium sources as well as the carbon and nitrogen cycling processes affecting ammonium fate. Specifically, RFC was used to unravel the different key factors controlling the high levels of ammonium prevailing in the shallow and deep aquifer sections, respectively. Our results reveal that high concentrations of ammonium in the shallow groundwater system primarily originate from anthropogenic sources, before being modulated by intensive microbially mediated nitrogen transformation processes such as nitrification, denitrification and dissimilatory nitrate reduction to ammonium (DNRA). By contrast, the linkage between high concentrations of ammonium and decomposition of soil organic matter, which ubiquitously contained nitrogen, suggested that mineralization of soil organic nitrogen compounds is the primary mechanism for the enrichment of ammonium in deeper groundwaters.

Drinking water sources and water quality in a prospective agricultural cohort

We describe drinking water sources and water quality for a large agricultural cohort. We used questionnaire data from the Agricultural Health Study (N = 89,655), a cohort of licensed pesticide applicators and their spouses in Iowa (IA) and North Carolina (NC), to ascertain drinking water source at enrollment (1993-1997). For users of public water supplies (PWS), we linked participants' geocoded addresses to contaminant monitoring data [five haloacetic acids (HAA5), total trihalomethanes (TTHM), and nitrate-nitrogen (NO3-N)]. We estimated private well nitrate levels using random forest models accounting for well depth, soil characteristics, nitrogen inputs, and other predictors. We assigned drinking water source for 84% (N = 74,919) of participants. Among these, 69% of IA and 75% of NC participants used private wells; 27% in IA and 21% in NC used PWS. Median PWS nitrate concentrations (NO3-N) were higher in IA [0.9 mg/L, interquartile range (IQR): 0.4-3.1 mg/L] than NC (0.1 mg/L, IQR: 0.1-0.2 mg/L), while median HAA5 and TTHM concentrations were higher in NC (HAA5: 11.9 µg/L, IQR: 5.5-33.4 µg/L; TTHM: 37.7 µg/L, IQR: 10.7-54.7 µg/L) than IA (HAA5: 5.0 µg/L, IQR: 3.7-10.7 µg/L; TTHM: 13.0 µg/L, IQR: 4.2-32.4 µg/L). Private well nitrate concentrations in IA (1.5 mg/L, IQR: 0.8-4.9 mg/L) and NC (1.9 mg/L, IQR: 1.4-2.5 mg/L) were higher than PWS. More private wells in IA (12%) exceeded 10 mg/L NO3-N (regulatory limit for PWS) than NC (<1%). Due to the proximity of their drinking water sources to farms, agricultural communities may be exposed to elevated nitrate levels.

Examining Relationships Between Groundwater Nitrate Concentrations in Drinking Water and Landscape Characteristics to Understand Health Risks

Nitrate ingested from drinking water has been linked to adverse health outcomes (e.g., cancer, birth defects) at levels as low as ∼2 mg/L NO₃-N, far below the regulatory limits of 10 mg/L. In many areas, groundwater is a common drinking water source and may contain elevated nitrate, but limited data on the patterns and concentrations are available. Using an extensive regulatory data set of over 100,000 nitrate drinking water well samples, we developed new maps of groundwater nitrate concentrations from 76,724 wells in Michigan's Lower Peninsula, USA for the 2006-2015 period. Kriging, a geostatistical method, was used to interpolate concentrations and quantify probability of exceeding relevant thresholds (>0.4 [common detection limit], >2 mg/L NO₃-N). We summarized this probability in small watersheds (∼80 km²) to identify correlated variables using the machine learning method classification and regression trees (CARTs). We found 79% of wells had concentrations below the detection limit in this analysis (<0.4 mg/L NO₃-N). In the shallow aquifer (focus of study), 13% of wells exceeded 2 mg/L NO₃-N and 2% exceeded the EPA maximum contaminant level of 10 mg/L. CART explained 40%-45% of variation in each model and identified three categories of critical correlated variables: source (high agricultural nitrogen inputs), vulnerable soil conditions (low soil organic carbon and high hydraulic conductivity), and transport mechanisms (high aquifer recharge). These findings add to the body of literature seeking to identify communities at risk of elevated nitrate and study associated adverse health outcomes.

Drinking Water Disinfection Byproducts, Ingested Nitrate, and Risk of Endometrial Cancer in Postmenopausal Women

BACKGROUND

Disinfection byproducts (DBPs) and N-nitroso compounds (NOC), formed endogenously after nitrate ingestion, are suspected endometrial carcinogens, but epidemiological studies are limited.

OBJECTIVES

We investigated the relationship of these exposures with endometrial cancer risk in a large prospective cohort.

METHODS

Among postmenopausal women in the Iowa Women's Health Study cohort, we evaluated two major classes of DBPs, total trihalomethanes (TTHM) and five haloacetic acids (HAA5), and nitrate-nitrogen (NO3-N) in public water supplies (PWS) in relation to incident primary endometrial cancer (1986-2014). For women using their PWS >10y at enrollment (n=10,501; cases=261), we computed historical averages of annual concentrations; exposures were categorized into quantiles and when possible ≥95th percentile. We also computed years of PWS use above one-half the U.S. maximum contaminant level (>½ MCL; 40μg/L TTHM; 30μg/L HAA5; 5mg/L NO3-N). Dietary nitrate/nitrite intakes were estimated from a food frequency questionnaire. We estimated hazard ratios (HR) and 95% confidence intervals (CI) via Cox models adjusted for age, endometrial cancer risk factors [e.g., body mass index, hormone replacement therapy (HRT)], and mutually adjusted for DBPs or NO3-N. We evaluated associations for low-grade (cases=99) vs. high-grade (cases=114) type I tumors. We assessed interactions between exposures and endometrial cancer risk factors and dietary factors influencing NOC formation.

RESULTS

Higher average concentrations of DBPs (95th percentile: TTHM ≥93μg/L, HAA5 ≥49μg/L) were associated with endometrial cancer risk (TTHM: HR95vsQ1=2.19, 95% CI: 1.41, 3.40; HAA5: HR95vsQ1=1.84, 95% CI: 1.19, 2.83; ptrend<0.01). Associations were similarly observed for women greater than median years of PWS use with levels >½ MCL, in comparison with zero years (TTHM: HR36+vs0y=1.61, 95% CI: 1.18, 2.21; HAA5: HR38+vs0y=1.85, 95% CI: 1.31, 2.62). Associations with DBPs appeared stronger for low-grade tumors (TTHM: HRQ4vsQ1=2.12, 95% CI: 1.17, 3.83; p-trend=0.008) than for high-grade tumors (TTHM: HRQ4vsQ1=1.40, 95% CI: 0.80, 2.44; p-trend=0.339), but differences were not statistically significant (p-heterogeneity=0.43). Associations with TTHM were stronger among ever HRT users than non-HRT users (p-interaction<0.01). We observed no associations with NO3-N in drinking water or diet.

DISCUSSION

We report novel associations between the highest DBP levels and endometrial cancer for our Iowa cohort that warrant future evaluation. https://doi.org/10.1289/EHP10207.

Sulfidation of ferric (hydr)oxides and its implication on contaminants transformation: a review

Rapid industrialization and urbanization have resulted in elevated concentrations of contaminants in the groundwaters and subsurface soils, posing a growing hazard to humans and ecosystems. The transformation of most contaminants is closely linked to the mineralogy of ferric (hydr)oxides. Sulfidation of ferric (hydr)oxides is one of the most significant biogeochemical reactions in the anoxic environments, causing reductive dissolution and recrystallization of ferric (hydr)oxides and further affecting the transformation of iron-associated contaminants. This paper provides a comprehensive review on the sulfidation process of ferric (hydr)oxides and the transformation of relevant contaminants. This review presents detailed reaction mechanisms between ferric (hydr)oxides and dissolved sulfide, and elucidates the factors (e.g. crystallinity of ferric (hydr)oxides, the ratio of sulfide concentration to the surface area concentration of ferric (hydr)oxides) that control the formation of surface associated Fe(II), iron sulfide minerals, as well as transformation of secondary minerals. Then, we summarized the transformation mechanisms of a variety of typical environmentally relevant contaminants existing in groundwater and subsurface soils, including heavy metals, metal(loid) oxyanions (arsenic, antimony, chromium), radionuclides (uranium, technetium), organic contaminants and phosphate/nitrate species. The general mechanisms of contaminant transformation involve a combination of release, reduction and re-adsorption/incorporation processes, the specific pathway of which is highly dependent on the properties of the contaminant itself and the extent of sulfidation. Moreover, the challenge of extending our knowledge towards in situ remediation, as well as further research needs are identified.

Predictive modeling of groundwater nitrate pollution and evaluating its main impact factors using random forest

Groundwater quality in plains and basins of arid and semi-arid regions with increased agriculture and urbanization development faces severe nitrate pollution, which is affected by both climate and anthropogenic activities. Here, shallow groundwater nitrate concentrations in the Yinchuan Region in central Yinchuan Plain were modeled during 2000, 2005, 2010, and 2015 using random forest. Multiple spatial environment factors were taken as predictor variables. The relative importance of these factors was also calculated using the constructed model. Remote sensing and GIS methods were used to compile various environmental factors to generate training and test sets for training and validation of the random forest model. Mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²) between the observed and predicted groundwater nitrate concentrations were used to measure the model performance. As indicated by these metrics, the random forest model for groundwater nitrate prediction was performed well. The relative importance of the predictor variables computed by the model indicated groundwater nitrate was mainly affected by the distance to the Yellow River, meteorological elements (precipitation, evaporation, and mean air temperature), and water level elevation. Additionally, urban and arable land were the two land use/land cover types that mainly influenced groundwater nitrate concentration in the Yinchuan Region, of which urban land was more influential than arable land as a result of intense expansion of urban land from 2000 to 2015. Overall, the current study provides an approach to integrate multiple environmental factors for groundwater quality study and is also significant for sustainable groundwater management in the Yinchuan Region.

Prediction of groundwater nitrate concentration in a semiarid region using hybrid Bayesian artificial intelligence approaches

Nitrate is a major pollutant in groundwater whose main source is municipal wastewater and agricultural activities. In the present study, Bayesian approaches such as Bayesian generalized linear model (BGLM), Bayesian regularized neural network (BRNN), Bayesian additive regression tree (BART), and Bayesian ridge regression (BRR) were used to model groundwater nitrate contamination in a semiarid region Marvdasht watershed, Fars province, Iran. Eleven groundwater (GW) nitrate conditioning factors have been taken as input parameters for predictive modeling. The results showed that the Bayesian models used in this study were all competent to model groundwater nitrate and the BART model with R² = 0.83 was more efficient than the other models. The result of variable importance showed that potassium (K) has the highest importance in the models followed by rainfall, altitude, groundwater depth, and distance from the residential area. The results of the study can support the decision-making process to control and reduce the sources of nitrate pollution.

Machine learning predictions of nitrate in groundwater used for drinking supply in the conterminous United States

Groundwater is an important source of drinking water supplies in the conterminous United State (CONUS), and presence of high nitrate concentrations may limit usability of groundwater in some areas because of the potential negative health effects. Prediction of locations of high nitrate groundwater is needed to focus mitigation and relief efforts. A three-dimensional extreme gradient boosting (XGB) machine learning model was developed to predict the distribution of nitrate. Nitrate was predicted at a 1 km resolution for two drinking water zones, each of variable depth, one for domestic supply and one for public supply. The model used measured nitrate concentrations from 12,082 wells and included predictor variables representing well characteristics, hydrologic conditions, soil type, geology, land use, climate, and nitrogen inputs. Predictor variables derived from empirical or numerical process-based models were also included to integrate information on controlling processes and conditions. The model provided accurate estimates at national and regional scales: the training (R² of 0.83) and hold-out (R² of 0.49) data fits compared favorably to previous studies. Predicted nitrate concentrations were less than 1 mg/L across most of the CONUS. Nationally, well depth, soil and climate characteristics, and the absence of developed land use were among the most influential explanatory factors. Only 1% of the area in either water supply zone had predicted nitrate concentrations greater than 10 mg/L; however, about 1.4 M people depend on groundwater for their drinking supplies in those areas. Predicted high concentrations of nitrate were most prevalent in the central CONUS. In areas of predicted high nitrate concentration, applied manure, farm fertilizer, and agricultural land use were influential predictor variables. This work represents the first application of XGB to a three-dimensional national-scale groundwater quality model and provides a significant milestone in the efforts to document nitrate in groundwater across the CONUS.

Identifying schools at high-risk for elevated lead in drinking water using only publicly available data

Estimating the risk of lead contamination of schools' drinking water at the State level is a complex, important, and unexplored challenge. Variable water quality among water systems and changes in water chemistry during distribution affect lead dissolution rates from pipes and fittings. In addition, the locations of lead-bearing plumbing materials are uncertain. We tested the capability of six machine learning models to predict the likelihood of lead contamination of drinking water at the schools' taps using only publicly available datasets. The predictive features used in the models correspond to those with a proven correlation to the dominant, but commonly unavailable, factors that govern lead leaching: the presence of lead-bearing plumbing materials and water quality conducive to lead corrosion. By combining water chemistry data from public reports, socioeconomic information from the US census, and spatial features using Geographic Information Systems, we trained and tested models to estimate the likelihood of lead contaminated tap water in over 8,000 schools across California and Massachusetts. Our best-performing model was a Random Forest, with a 10-fold cross validation score of 0.88 for Massachusetts and 0.78 for California using the average Area Under the Receiver Operating Characteristic Curve (ROC AUC) metric. The model was then used to assign a lead leaching risk category to half of the schools across California (the other half was used for training). There was good agreement between the modeled risk categories and the actual lead leaching outcomes for every school; however, the model overestimated the lead leaching risk in up to 17% of the schools. This model is the first of its kind to offer a tool to predict the risk of lead leaching in schools at the State level. Further use of this model can help deploy limited resources more effectively to prevent childhood lead exposure from school drinking water.

Electrical conductivity and dissolved oxygen as predictors of nitrate concentrations in shallow groundwater in Erhai Lake region

Elevated nitrogen (N) concentration in shallow groundwater is becoming increasingly problematic, putting water resources under pressure. For more effective management of such a resource, more precise predictors of N level in groundwater using smart monitoring networks are needed. However, external factors such as land use type, rainfall, and N loads from multiple sources (residential and agricultural) make it difficult to accurately predict the spatial and temporal variations of N concentration. In order to identify the key factors affecting spatial and temporal N concentration in shallow groundwater and develop a predictive model, 635 groundwater samples from drinking wells in residential areas and agricultural wells in croplands of a typical agricultural watershed in the Erhai Lake Basin, southwest China, in the period from 2018 to 2020, were collected and analyzed. The results showed that the type of land use and seasonal variations significantly affected the N forms and their concentrations in the shallow groundwater, as the ratios of ON and NO₃^--N to TN were 30%-39% and 52%-59% for the two land uses and 25%-44% and 46%-66% for seasonal changes. Their variations were reflected by electrical conductivity (EC) and redox environment. EC and dissolved oxygen (DO) had a positive non-linear relationship with the concentrations of total nitrogen (TN) and nitrate (NO₃^--N). The fitted non-linear quantitative models were established separately to predict TN and NO₃^--N concentrations in groundwater using easily available indictors (EC and DO). The high accuracy and performance of the models were investigated and approved by rRMSE, MAE, and 1:1 line. These findings can provide technical support for the rapid prediction and evaluation of N pollution in shallow groundwater through easily available indicators.

Development of a Mechanically Flexible 2D-MXene Membrane Cathode for Selective Electrochemical Reduction of Nitrate to N2: Mechanisms and Implications

The contamination of water resources by nitrate is a major problem. Herein, we report a mechanically flexible 2D-MXene (Ti₃C₂T_x) membrane with multilayered nanofluidic channels for a selective electrochemical reduction of nitrate to nitrogen gas (N₂). At a low applied potential of -0.8 V (vs Ag/AgCl), the MXene electrochemical membrane was found to exhibit high selectivity for NO₃^- reduction to N₂ (82.8%) due to a relatively low desorption energy barrier for the release of adsorbed N₂ (*N₂) compared to that for the adsorbed NH₃ (*NH₃) based on density functional theory (DFT) calculations. Long-term use of the MXene membrane for treating 10 mg-NO₃-N L^-1 in water was found to have a high faradic efficiency of 72.6% for NO₃^- reduction to N₂ at a very low electrical cost of 0.28 kWh m^-3. Results of theoretical calculations and experimental results showed that defects on the MXene nanosheet surfaces played an important role in achieving high activity, primarily at the low-coordinated Ti sites. Water flowing through the MXene nanosheets facilitated the mass transfer of nitrate onto the low-coordinated Ti sites with this enhancement of particular importance under cathodic polarization of the MXene membrane. This study provides insight into the tailoring of nanoengineered materials for practical application in water treatment and environmental remediation.

Ingestion of Nitrate and Nitrite and Risk of Stomach and Other Digestive System Cancers in the Iowa Women's Health Study

Nitrate and nitrite are precursors in the endogenous formation of N-nitroso compounds (NOC) which are potent animal carcinogens for the organs of the digestive system. We evaluated dietary intakes of nitrate and nitrite, as well as nitrate ingestion from drinking water (public drinking water supplies (PWS)), in relation to the incidence (1986-2014) of cancers of the esophagus (n = 36), stomach (n = 84), small intestine (n = 32), liver (n = 31), gallbladder (n = 66), and bile duct (n = 58) in the Iowa Women's Health Study (42,000 women aged from 50 to 75 in 1986). Dietary nitrate and nitrite were estimated using a food frequency questionnaire and a database of nitrate and nitrite levels in foods. Historical nitrate measurements from PWS were linked to the enrollment address by duration. We used Cox regression to compute hazard ratios (HR) and 95% confidence intervals (CI) for exposure quartiles (Q), tertiles (T), or medians, depending on the number of cancer cases. In adjusted models, nitrite intake from processed meats was associated with an increased risk of stomach cancer (HRQ4vsQ1 = 2.2, CI: 1.2-4.3). A high intake of total dietary nitrite was inversely associated with gallbladder cancer (HRQ4vsQ1 = 0.3, CI: 0.1-0.96), driven by an inverse association with plant sources of nitrite (HRQ4vsQ1 = 0.3, CI: 0.1-0.9). Additionally, small intestine cancer was inversely associated with a high intake of animal nitrite (HRT3vsT1 = 0.2, CI: 0.1-0.7). There were no other dietary associations. Nitrate concentrations in PWS (average, years ≥ 1/2 the maximum contaminant level) were not associated with cancer incidence. Our findings for stomach cancer are consistent with prior dietary studies, and we are the first to evaluate nitrate and nitrite ingestion for certain gastrointestinal cancers.

Sources and Risk Factors for Nitrate and Microbial Contamination of Private Household Wells in the Fractured Dolomite Aquifer of Northeastern Wisconsin

BACKGROUND

Groundwater quality in the Silurian dolomite aquifer in northeastern Wisconsin, USA, has become contentious as dairy farms and exurban development expand.

OBJECTIVES

We investigated private household wells in the region, determining the extent, sources, and risk factors of nitrate and microbial contamination.

METHODS

Total coliforms, Escherichia coli, and nitrate were evaluated by synoptic sampling during groundwater recharge and no-recharge periods. Additional seasonal sampling measured genetic markers of human and bovine fecal-associated microbes and enteric zoonotic pathogens. We constructed multivariable regression models of detection probability (log-binomial) and concentration (gamma) for each contaminant to identify risk factors related to land use, precipitation, hydrogeology, and well construction.

RESULTS

Total coliforms and nitrate were strongly associated with depth-to-bedrock at well sites and nearby agricultural land use, but not septic systems. Both human wastewater and cattle manure contributed to well contamination. Rotavirus group A, Cryptosporidium, and Salmonella were the most frequently detected pathogens. Wells positive for human fecal markers were associated with depth-to-groundwater and number of septic system drainfield within 229m. Manure-contaminated wells were associated with groundwater recharge and the area size of nearby agricultural land. Wells positive for any fecal-associated microbe, regardless of source, were associated with septic system density and manure storage proximity modified by bedrock depth. Well construction was generally not related to contamination, indicating land use, groundwater recharge, and bedrock depth were the most important risk factors.

DISCUSSION

These findings may inform policies to minimize contamination of the Silurian dolomite aquifer, a major water supply for the U.S. and Canadian Great Lakes region. https://doi.org/10.1289/EHP7813.

ANFIS-MOA models for the assessment of groundwater contamination vulnerability in a nitrate contaminated area

The enhanced assessment of groundwater contamination vulnerability is necessary for the management and conservation of groundwater resources because groundwater contamination has been much increased continuously in the world by anthropogenic origin. The purpose of this study is to determine the best model among three ANFIS-MOA models (the adaptive neuro-fuzzy inference system (ANFIS) combined with metaheuristic optimization algorithms (MOAs) such as genetic algorithm (GA), differential evolution algorithm (DE) and particle swarm optimization (PSO)) in assessing groundwater contamination vulnerability at a nitrate contaminated area. The Miryang City of South Korea was selected as the study area because the nitrate contamination was widespread in the city with two functions of urban and rural activities. Eight parameters (depth to water, net recharge, topographic slope, aquifer type, impact to vadose zone, hydraulic conductivity and landuse) were classified into the numerical ratings on basis of modified DRASTIC method (MDM) for the input variables of ANFIS-MOA models. The Original ANFIS, and 3 combined models of ANFIS-PSO, ANFIS-DE and, ANFIS-GA used 95 adjusted vulnerability indices (AVI) as the target data of training (70% data) and testing (30% data) processing. The performance of 4 models was evaluated by mean absolute errors (MAE), root mean square errors (RMSE), correlation coefficients (R), ROC/AUC curves and predicted AVI (PAVI) maps. The statistical results, spatial vulnerability maps and correlation coefficients between PAVIs and nitrate concentrations revealed that the order of model excellence was ANFIS-PSO, ANFIS-DE, ANFIS-GA, and Original ANFIS, and that ANFIS-PSO showed the highest performance in training and testing processing. The performance rates of ANFIS-MOA models were also compared with 10 recent popular worldwide models using the correlation coefficients between PVI and nitrate concentrations, and they were superior to other recent popular models. ANFIS-MOA models were also useful for resolving the subjectivity of physical and hydrogeological parameters in original DRASTIC method (ODM) and MDM. It is expected that ANFIS-PSO models will produce the excellent results in assessing groundwater contamination vulnerability and that they can greatly contribute to the groundwater security in other areas of the world as well as Miryang City of South Korea.

Modeling elevated blood lead level risk across the United States

Lead exposure adversely affects child health and continues to be a major public health concern in the United States (US). Lead exposure risk has been linked with older housing and households in poverty, but more studies of neighborhood socioeconomic status (SES) and lead exposure risk over large and diverse geographic areas are needed. In this paper, we combined lead test result data over many states for a majority of the US ZIP Codes in order to estimate its association with many SES variables and predict lead exposure risk in all populated ZIP Codes in the US. The methods used for estimation and prediction of lead risk included the Vox lead exposure risk score, random forest, weighted quantile sum (WQS) regression, and a Bayesian SES index model. The results showed that the Bayesian index model had the best overall performance for modeling elevated blood lead level (EBLL) risk and therefore was used to create a lead exposure risk score for US ZIP Codes. There was a statistically significant association between EBLL risk and the SES index and the most important SES variables for explaining EBLL risk were percentage of houses built before 1940 and median home value. When mapping the lead exposure risk scores, there was a clear pattern of elevated risk in the Northeast and Midwest, but areas in the South and Southwest regions of the US also had high risk. In summary, the Bayesian index model was an effective method for modeling EBLL risk associated with neighborhood deprivation while accounting for additional heterogeneity in risk using lead test result data covering a majority of the US. The resulting lead exposure risk score can be used for targeting public health intervention efforts.

Machine Learning Predictions of pH in the Glacial Aquifer System, Northern USA

A boosted regression tree model was developed to predict pH conditions in three dimensions throughout the glacial aquifer system of the contiguous United States using pH measurements in samples from 18,386 wells and predictor variables that represent aspects of the hydrogeologic setting. Model results indicate that the carbonate content of soils and aquifer materials strongly controls pH and, when coupled with long flowpaths, results in the most alkaline conditions. Conversely, in areas where glacial sediments are thin and carbonate-poor, pH conditions remain acidic. At depths typical of drinking-water supplies, predicted pH >7.5-which is associated with arsenic mobilization-occurs more frequently than predicted pH <6-which is associated with water corrosivity and the mobilization of other trace elements. A novel aspect of this model was the inclusion of numerically based estimates of groundwater flow characteristics (age and flowpath length) as predictor variables. The sensitivity of pH predictions to these variables was consistent with hydrologic understanding of groundwater flow systems and the geochemical evolution of groundwater quality. The model was not developed to provide precise estimates of pH at any given location. Rather, it can be used to more generally identify areas where contaminants may be mobilized into groundwater and where corrosivity issues may be of concern to prioritize areas for future groundwater monitoring.

Spatial Heterogeneity in Positional Errors: A Comparison of Two Residential Geocoding Efforts in the Agricultural Health Study

Geocoding is a powerful tool for environmental exposure assessments that rely on spatial databases. Geocoding processes, locators, and reference datasets have improved over time; however, improvements have not been well-characterized. Enrollment addresses for the Agricultural Health Study, a cohort of pesticide applicators and their spouses in Iowa (IA) and North Carolina (NC), were geocoded in 2012–2016 and then again in 2019. We calculated distances between geocodes in the two periods. For a subset, we computed positional errors using “gold standard” rooftop coordinates (IA; N = 3566) or Global Positioning Systems (GPS) (IA and NC; N = 1258) and compared errors between periods. We used linear regression to model the change in positional error between time periods (improvement) by rural status and population density, and we used spatial relative risk functions to identify areas with significant improvement. Median improvement between time periods in IA was 41 m (interquartile range, IQR: −2 to 168) and 9 m (IQR: −80 to 133) based on rooftop coordinates and GPS, respectively. Median improvement in NC was 42 m (IQR: −1 to 109 m) based on GPS. Positional error was greater in rural and low-density areas compared to in towns and more densely populated areas. Areas of significant improvement in accuracy were identified and mapped across both states. Our findings underscore the importance of evaluating determinants and spatial distributions of errors in geocodes used in environmental epidemiology studies.

Private Wells and Rural Health: Groundwater Contaminants of Emerging Concern

PURPOSE OF REVIEW

Approximately 12% of the population in the US and Canada rely on federally unregulated private wells, which are common in rural areas and may be susceptible to microbiological and chemical contamination. This review identifies and summarizes recent findings on contaminants of emerging concern in well water across the US and Canada.

RECENT FINDINGS

Private well water quality modeling is complicated by the substantial variability in contamination sources, well construction, well depth, and the hydrogeology of the environment surrounding the well. Temporal variation in contaminant levels in wells suggests the need for monitoring efforts with greater spatial and temporal coverage. More extensive private well monitoring will help identify wells at greater risk of contamination, and in turn, public health efforts can focus on education and outreach to improve monitoring, maintaining, and treating private wells in these communities. Community interventions need to be coupled with stricter regulations and financing mechanisms that can support and protect private well owners.

Soil and Aquifer Properties Combine as Predictors of Groundwater Uranium Concentrations within the Central Valley, California

With the increasing global need for groundwater resources to fulfill domestic, agricultural, and industrial demands, we face the threat of increasing concentrations of naturally occurring contaminants in water sources and a consequential need to improve our predictive capacity. Here, we combine machine learning and geochemical modeling to reveal the biogeochemical controls on regional groundwater uranium contamination within the Central Valley, California. We use 23 environmental parameters from a statewide groundwater geochemical database and publicly available maps of soil and aquifer physicochemical properties to predict groundwater uranium concentrations by random forest regression. We find that groundwater calcium, nitrate, and sulfate concentrations, soil pH, and clay content (weighted average between 0 and 2 m depths) are the most important predictors of groundwater uranium concentrations. By pairing multivariate partial dependence and accumulated local effect plots with modeled aqueous uranium speciation and surface complexation outputs, we show that regional groundwater uranium exceedances of drinking water standards, 30 μg L^-1, are dependent on the formation of uranyl-calcium-carbonato species. The geochemical conditions leading to ternary uranyl complexes within the aquifer are, in part, created by infiltration through the vadose zone, illustrating the critical dependence of groundwater quality on recharge conditions.

A Novel Approach to Harmonize Vulnerability Assessment in Carbonate and Detrital Aquifers at Basin Scale

The DRASTIC (D: Depth to water; R: Net recharge; A: Aquifer media; S: Soil media; T: Topography; I: Impact of vadose zone; C: Hydraulic conductivity) index is usually applied to assess intrinsic vulnerability in detrital and carbonate aquifers, although it does not take into account the particularities of karst systems as the COP (C: Concentration of flow; O: Overlying layers above water table; P: precipitation) method does. In this paper we aim to find a reasonable correspondence between the vulnerability maps obtained using these two methods. We adapt the DRASTIC index in order to obtain reliable assessments in carbonate aquifers while maintaining its original conceptual formulation. This approach is analogous to the hypothesis of “equivalent porous medium”, which applies to karstic aquifers the numerical solution developed for detrital aquifers. We applied our novel method to the Upper Guadiana Basin, which contains both carbonate and detrital aquifers. Validation analysis demonstrated a higher confidence in the vulnerability assessment provided by the COP method in the carbonate aquifers. The proposed method solves an optimization problem to minimize the differences between the assessments provided by the modified DRASTIC and COP methods. Decision trees and spatial statistics analyses were combined to identify the ranges and weights of DRASTIC parameters to produce an optimal solution that matches the COP vulnerability classification for carbonate aquifers in 75% of the area, while maintaining a reliable assessment of the detrital aquifers in the Basin.

Patterns and predictions of drinking water nitrate violations across the conterminous United States

Excess nitrate in drinking water is a human health concern, especially for young children. Public drinking water systems in violation of the 10 mg nitrate-N/L maximum contaminant level (MCL) must be reported in EPA's Safe Drinking Water Information System (SDWIS). We used SDWIS data with random forest modeling to examine the drivers of nitrate violations across the conterminous U.S. and to predict where public water systems are at risk of exceeding the nitrate MCL. As explanatory variables, we used land cover, nitrogen inputs, soil/hydrogeology, and climate variables. While we looked at the role of nitrate treatment in separate analyses, we did not include treatment as a factor in the final models, due to incomplete information in SDWIS. For groundwater (GW) systems, a classification model correctly classified 79% of catchments in violation and a regression model explained 43% of the variation in nitrate concentrations above the MCL. The most important variables in the GW classification model were % cropland, agricultural drainage, irrigation-to-precipitation ratio, nitrogen surplus, and surplus precipitation. Regions predicted to have risk for nitrate violations in GW were the Central California Valley, parts of Washington, Idaho, the Great Plains, Piedmont of Pennsylvania and Coastal Plains of Delaware, and regions of Wisconsin, Iowa, and Minnesota. For surface water (SW) systems, a classification model correctly classified 90% of catchments and a regression model explained 52% of the variation in nitrate concentration. The variables most important for the SW classification model were largely hydroclimatic variables including surplus precipitation, irrigation-to-precipitation ratio, and % shrubland. Areas at greatest risk for SW nitrate violations were generally in the non-mountainous west and southwest. Identifying the areas with possible risk for future violations and potential drivers of nitrate violations across U.S. can inform decisions on how source water protection and other management options could best protect drinking water.

Enhancing nitrate and strontium concentration prediction in groundwater by using new data mining algorithm

Groundwater resources constitute the main source of clean fresh water for domestic use and it is essential for food production in the agricultural sector. Groundwater has a vital role for water supply in the Campanian Plain in Italy and hence a future sustainability of the resource is essential for the region. In the current paper novel data mining algorithms including Gaussian Process (GP) were used in a large groundwater quality database to predict nitrate (contaminant) and strontium (potential future increasing) concentrations in groundwater. The results were compared with M5P, random forest (RF) and random tree (RT) algorithms as a benchmark to test the robustness of the modeling process. The dataset includes 246 groundwater quality samples originating from different wells, municipals and agricultural. It was divided for the modeling process into two subgroups by using the 10-fold cross validation technique including 173 samples for model building (training dataset) and 73 samples for model validation (testing dataset). Different water quality variables including T, pH, EC, HCO₃^-, F^-, Cl^-, SO₄^2-, Na⁺, K⁺, Mg²⁺, and Ca²⁺ have been used as an input to the models. At first stage, different input combinations have been constructed based on correlation coefficient and thus the optimal combination was chosen for the modeling phase. Different quantitative criteria alongside with visual comparison approach have been used for evaluating the modeling capability. Results revealed that to obtain reliable results also variables with low correlation should be considered as an input to the models together with those variables showing high correlation coefficients. According to the model evaluation criteria, GP algorithm outperforms all the other models in predicting both nitrate and strontium concentrations followed by RF, M5P and RT, respectively. Result also revealed that model's structure together with the accuracy and structure of the data can have a relevant impact on the model's results.

Predicting uncertainty of machine learning models for modelling nitrate pollution of groundwater using quantile regression and UNEEC methods

Although estimating the uncertainty of models used for modelling nitrate contamination of groundwater is essential in groundwater management, it has been generally ignored. This issue motivates this research to explore the predictive uncertainty of machine-learning (ML) models in this field of study using two different residuals uncertainty methods: quantile regression (QR) and uncertainty estimation based on local errors and clustering (UNEEC). Prediction-interval coverage probability (PICP), the most important of the statistical measures of uncertainty, was used to evaluate uncertainty. Additionally, three state-of-the-art ML models including support vector machine (SVM), random forest (RF), and k-nearest neighbor (kNN) were selected to spatially model groundwater nitrate concentrations. The models were calibrated with nitrate concentrations from 80 wells (70% of the data) and then validated with nitrate concentrations from 34 wells (30% of the data). Both uncertainty and predictive performance criteria should be considered when comparing and selecting the best model. Results highlight that the kNN model is the best model because not only did it have the lowest uncertainty based on the PICP statistic in both the QR (0.94) and the UNEEC (in all clusters, 0.85-0.91) methods, but it also had predictive performance statistics (RMSE = 10.63, R² = 0.71) that were relatively similar to RF (RMSE = 10.41, R² = 0.72) and higher than SVM (RMSE = 13.28, R² = 0.58). Determining the uncertainty of ML models used for spatially modelling groundwater-nitrate pollution enables managers to achieve better risk-based decision making and consequently increases the reliability and credibility of groundwater-nitrate predictions.

Impact of land cover on groundwater quality in the Upper Floridan Aquifer in Florida, United States

Although agricultural lands are generally assumed to correlate negatively with groundwater quality, the intricate relationship between general land cover and contaminant concentrations present in an aquifer may vary substantially; contingent upon the land type, interacting factors, and scale considered. The Upper Floridan Aquifer (UFA) is a primary source of potable water supply for the state of Florida. The Suwannee River Water Management District (SRWMD), located in northcentral Florida, relies exclusively on the UFA for water supplies. Over much of the SRWMD in the UFA is unconfined, rendering it vulnerable to contamination from surface sources. This study analyses groundwater concentrations of Nitrate-Nitrogen (NO₃-N) and Potassium (K) from shallow wells across the SRWMD for assessing the effect of different land covers on groundwater quality over time. Annual potentiometric surface maps were used to delineate semicircular recharge zones of 500 m, 1000 m, and 2000 m radii upstream of sampled well stations. Proportions of agriculture, forest, and urban lands were identified for each buffer zone using USDA Cropland Data Layer. Multivariate regression models were developed to infer relationships between land cover and NO₃-N and K concentrations. Results show significant associations among land cover type, water table height, and groundwater quality parameters. Specifically, we find a large proportion of agricultural cover consistently associated with larger increases in groundwater pollutant loads relative to urban or forest cover across all models, after controlling for depth to water table. Our study suggests a need for widespread adoption of cost-effective agricultural best management practices (BMPs) that could help in securing regional water supply.

Large scale prediction of groundwater nitrate concentrations from spatial data using machine learning

Reducing nitrogen inputs, in particular nitrate, to groundwater is becoming increasingly important to fulfil requirements of the European Water Framework Directive. When developing management plans for mitigation measures at larger scales, complex hydro-biogeochemical models reach their limits due to data availability and spatial discretization. To circumvent this problem, the spatial distribution of nitrate concentration in groundwater is estimated using a parsimonious GIS-based statistical approach. Point nitrate concentrations and spatial environmental data as predictors are used to train statistical models. In order to compile the spatial predictors with the respective monitoring sites, different designs of contributing areas (buffer zones) and their effects on the performance of different statistical models are investigated. Multiple Linear Regression (MLR), Classification and Regression Trees (CART), Random Forest (RF) and Boosted Regression Trees (BRT) are compared in terms of the predictive performance of each model according to various objective functions. We determine the most influential spatial predictors used in the respective models. After training the models with a subset of the data, we then predict the spatial nitrate distribution in groundwater for the entire federal state of Hesse, Germany on a 1 × 1 km grid by only the spatial environmental data. The Random Forest model outperforms the other models (R² = 0.54), relying on hydrogeological units, the percentage of arable land and the nitrogen balance as the three most influencing predictors based on a 1000 m circular contributing area. The use of exclusively spatial available predictors is a big step forward in the prediction of nitrate in groundwater on regional scale.

Operation of a 2-Stage Bioelectrochemical System for Groundwater Denitrification

Nitrate groundwater contamination is an issue of global concern that has not been satisfactorily and efficiently addressed, yet. In this study, a 2-stage, sequential bioelectrochemical system (BES) was run to perform autotrophic denitrification of synthetic groundwater. The system was run at a 75.6 mgNO3−-N L−1NCC d−1 nitrate loading rate, achieving almost complete removal of nitrate (>93%) and Total Nitrogen (TN) (>93%). After treatment in the first stage reactor values of effluent nitrate compatible with the EU and USA limits for drinking water (<11.3 and 10 mgNO3−-N L−1, respectively) were achieved. Nitrite and nitrous oxide were observed in the first stage’s effluent, and were then successfully removed in the second stage. The observed nitrate removal rate was 73.4 ± 1.3 gNO3−-N m−3NCC d−1, while the total nitrogen removal rate was 73.1 ± 1.2 gN m−3NCC d−1. Specific energy consumptions of the system were 0.80 ± 0.00 kWh m−3, 18.80 ± 0.94 kWh kgNO3−-N−1 and 18.88 ± 0.95 kWh kgN−1. Combination of two denitrifying BES in series herein described proved to be effective.

Ingested nitrate, disinfection by-products, and risk of colon and rectal cancers in the Iowa Women's Health Study cohort

BACKGROUND

N-nitroso compounds (NOC) formed endogenously after nitrate/nitrite ingestion and disinfection by-products (DBPs) are suspected colorectal carcinogens, but epidemiologic evidence of these associations is limited.

OBJECTIVES

We investigated the relationship between drinking water exposures and incident colorectal cancers in a cohort of postmenopausal women.

METHODS

Using historical nitrate-nitrogen (NO₃-N) measurements and estimates of total trihalomethanes (TTHM), the sum of 5 or 6 haloacetic acids (HAAs), and individual DBPs in public water supplies (PWS), we computed average exposures and years of exposure above one-half the U.S. maximum contaminant level (>1/2-MCL; >5 mg/L NO₃-N and >40 μg/L TTHM). Nitrate/nitrite intakes from dietary sources were estimated using a food frequency questionnaire. We estimated hazard ratios (HR) and 95% confidence intervals (CI) from Cox regression models. We assessed NO₃-N interactions with DBPs and with factors influencing endogenous NOC formation.

RESULTS

We identified 624 colon and 158 rectal cancers (1986-2010) among 15,910 women reporting PWS use >10 years. Ingestion of NO₃-N from drinking water was not associated with risk. Colon cancer risks were non-significantly associated with the average TTHM levels >17.7 μg/L (HR_Q5vsQ1 = 1.13, CI = 0.89-1.44; p_trend = 0.11) and were elevated for any duration of exposure >1/2-MCL. Rectal cancer risks were associated with the highest TTHM levels (HR_Q5vsQ1 = 1.71, CI = 1.00-2.92; p_trend = 0.22) but not with years >1/2-MCL. Bromodichloromethane (HR_Q4vsQ1 = 1.89, CI = 1.17-3.00; p_trend = 0.09) and trichloroacetic acid (HR_Q4vsQ1 = 1.92, CI = 1.20-3.09; p_trend = 0.18) levels were also associated with risk of rectal cancer. We found no evidence of interaction between TTHM and NO₃-N on the risk of either cancer. Dietary analyses yielded a positive colon cancer association with red meat, but not with processed meat intake or estimated nitrate/nitrite from specific dietary sources.

CONCLUSIONS

Our results suggest that exposure to TTHM in drinking water is associated with increased risk of rectal cancer. Positive findings for individual THMs and HAAs for both colon and rectal cancers require replication in other studies. We found no associations for nitrate overall or in subgroups with presumed higher NOC exposure.

Modeling groundwater nitrate exposure in private wells of North Carolina for the Agricultural Health Study

Unregulated private wells in the United States are susceptible to many groundwater contaminants. Ingestion of nitrate, the most common anthropogenic private well contaminant in the United States, can lead to the endogenous formation of N-nitroso-compounds, which are known human carcinogens. In this study, we expand upon previous efforts to model private well groundwater nitrate concentration in North Carolina by developing multiple machine learning models and testing against out-of-sample prediction. Our purpose was to develop exposure estimates in unmonitored areas for use in the Agricultural Health Study (AHS) cohort. Using approximately 22,000 private well nitrate measurements in North Carolina, we trained and tested continuous models including a censored maximum likelihood-based linear model, random forest, gradient boosted machine, support vector machine, neural networks, and kriging. Continuous nitrate models had low predictive performance (R2 < 0.33), so multiple random forest classification models were also trained and tested. The final classification approach predicted <1 mg/L, 1-5 mg/L, and ≥5 mg/L using a random forest model with 58 variables and maximizing the Cohen's kappa statistic. The final model had an overall accuracy of 0.75 and high specificity for the higher two categories and high sensitivity for the lowest category. The results will be used for the categorical prediction of private well nitrate for AHS cohort participants that reside in North Carolina.

Validating a continental-scale groundwater diffuse pollution model using regional datasets

In this study, we assess the validity of an African-scale groundwater pollution model for nitrates. In a previous study, we identified a statistical continental-scale groundwater pollution model for nitrate. The model was identified using a pan-African meta-analysis of available nitrate groundwater pollution studies. The model was implemented in both Random Forest (RF) and multiple regression formats. For both approaches, we collected as predictors a comprehensive GIS database of 13 spatial attributes, related to land use, soil type, hydrogeology, topography, climatology, region typology, nitrogen fertiliser application rate, and population density. In this paper, we validate the continental-scale model of groundwater contamination by using a nitrate measurement dataset from three African countries. We discuss the issue of data availability, and quality and scale issues, as challenges in validation. Notwithstanding that the modelling procedure exhibited very good success using a continental-scale dataset (e.g. R² = 0.97 in the RF format using a cross-validation approach), the continental-scale model could not be used without recalibration to predict nitrate pollution at the country scale using regional data. In addition, when recalibrating the model using country-scale datasets, the order of model exploratory factors changes. This suggests that the structure and the parameters of a statistical spatially distributed groundwater degradation model for the African continent are strongly scale dependent.

Evaluation of energy consumption of treating nitrate-contaminated groundwater by bioelectrochemical systems

Nitrate contamination of groundwater is a mounting concern for drinking water production due to its healthy and ecological effects. Bioelectrochemical systems (BES) are a promising method for energy efficient nitrate removal, but its energy consumption has not been well understood. Herein, we conducted a preliminary analysis of energy consumption based on both literature information and multiple assumptions. Four scenarios were created for the purpose of analysis based on two treatment approaches, microbial fuel cells (MFCs) and controlled biocathodic denitrification (CBD), under either in situ or ex situ deployment. The results show a specific energy consumption based on the mass of NO₃^--N removed (SEC_N) of 0.341 and 1.602 kWh kg NO₃^--N^-1 obtained from in situ and ex situ treatments with MFCs, respectively; the main contributor was the extraction of the anolyte (100%) in the former and pumping the groundwater (74.8%) for the latter. In the case of CBD treatment, the energy consumption by power supply outcompeted all the other energy items (over 85% in all cases), and a total SEC_N of 19.028 and 10.003 kWh kg NO₃^--N^-1 were obtained for in situ and ex situ treatments, respectively. The increase in the water table depth (from 10 to 30 m) and the decrease of the nitrate concentration (from 25 to 15 mg NO₃^--N) would lead to a rise in energy consumption in the ex situ treatment. Although some data might be premature due to the lack of sufficient information in available literature, the results could provide an initial picture of energy consumption by BES-based groundwater treatment and encourage further thinking and analysis of energy consumption (and production).