A generalized regression model of arsenic variations in the shallow groundwater of Bangladesh

Sulfate reduction accelerates groundwater arsenic contamination even in aquifers with abundant iron oxides

Groundwater contamination by geogenic arsenic is a global problem affecting nearly 200 million people. In South and Southeast Asia, a cost-effective mitigation strategy is to use oxidized low-arsenic aquifers rather than reduced high-arsenic aquifers. Aquifers with abundant oxidized iron minerals are presumably safeguarded against immediate arsenic contamination, due to strong sorption of arsenic onto iron minerals. However, preferential pumping of low-arsenic aquifers can destabilize the boundaries between these aquifers, pulling high-arsenic water into low-arsenic aquifers. We investigate this scenario in a hybrid field-column experiment in Bangladesh where naturally high-arsenic groundwater is pumped through sediment cores from a low-arsenic aquifer, and detailed aqueous and solid-phase measurements are used to constrain reactive transport modelling. Here we show that elevated groundwater arsenic concentrations are induced by sulfate reduction and the predicted formation of highly mobile, poorly sorbing thioarsenic species. This process suggests that contamination of currently pristine aquifers with arsenic can occur up to over 1.5 times faster than previously thought, leading to a deterioration of urgently needed water resources.

The Bengal Water Machine: Quantified freshwater capture in Bangladesh

Global food security depends on the sustainability of irrigated agriculture. Rising groundwater withdrawals from seasonally humid, alluvial plains across tropical Asia have enabled dry-season rice cultivation. This groundwater pumpage increases available subsurface storage that under favorable conditions amplifies groundwater replenishment during the subsequent monsoon. We empirically quantified this nature-based solution to seasonal freshwater storage capture described as the “Bengal Water Machine,” revealing its potential and limitations. On the basis of a million piezometric observations from 465 monitoring wells, we show that the collective operation of ~16 million smallholder farmers in the Bengal Basin of Bangladesh from 1988 to 2018 has induced cumulative freshwater capture that volumetrically (75 to 90 cubic kilometers) is equivalent to twice the reservoir capacity of the Three Gorges Dam.

The groundwater arsenic contamination in the Bengal Basin-A review in brief

The presence of arsenic in the groundwater of the densely-populated Bengal Basin evolved as a mass-poisoning agent and is a reason for the misery of millions of people living here. High-level arsenic was detected in the shallow aquifer-tube wells of the basin in the late-20th century. The redox conditions and the biogeochemical activities in the shallow aquifers support the existence of arsenic in its most toxic +3 state. The shallow aquifers are constructed by the Holocene reduced grey sands, having a lesser capacity to hold the arsenic brought from the Himalayas by the Ganga-Brahmaputra-Meghna river system. Among several other hypotheses, the reductive dissolution of arsenic bearing Fe-oxyhydroxides coupled with the microbial activities in the organic-matter-rich Holocene grey sands is believed to be the primary reason for releasing arsenic in groundwater of basinal shallow aquifers. The deep aquifers below the late Pleistocene aquifers and the Palaeo-interfluvial aquifers capped by the last glacial maximum Palaeosol generally contain arsenic-free or low-arsenic water. Ingress of arsenic into the deep aquifers from the shallow aquifers was considered to have been caused by extensive non-domestic pumping. However, studies have found that extensive pumping is unlikely to contaminate the deep aquifer water with higher levels of arsenic within decadal time scales. Irrigation-pumping may produce hydraulic barriers between the shallow and deep aquifer-groundwater and distributes arsenic in the topsoil by flushing. Significant disparities have been observed among the Bengal basinal groundwater arsenic concentrations. However, abrupt spatial variation in groundwater arsenic concentrations has been a key feature of the basin.

Increases in groundwater arsenic concentrations and risk under decadal groundwater withdrawal in the lower reaches of the Yellow River basin, Henan Province, China

The spatiotemporal variability in groundwater arsenic concentrations following extensive groundwater extractions over decades was rarely studied on a large scale. To fill this gap, variations in groundwater arsenic concentrations in the North Henan Plain in China from 2010 to 2020 were investigated. The possibility of high-arsenic groundwater (>10 μg/L) was higher than 40% in aquifers within a distance of 100 m from paleochannels. This may be due to the fact that deposits in paleochannels were rich in organic matter and suitable for arsenic enrichment. Following groundwater withdrawal over ten years from 2010 to 2020, nearly half of groundwater samples (44%) were elevated in groundwater arsenic concentrations, and the proportion of high arsenic groundwater increased from 24% in 2010 to 26% in 2020. These may be related to enhanced Fe(III) oxide reduction under decadal groundwater withdrawal. However, around 56% groundwater samples were decreases in arsenic concentrations because of increased NO₃^- levels in these samples in 2020. Furthermore, extensive groundwater withdrawal decreased groundwater tables averagely by 4.6 m from 2010 to 2020, which induced the intrusion of high-arsenic groundwater from shallow aquifers into deeper ones. More importantly, the long-term groundwater pumping has perturbed groundwater flow dynamics and redistributed high-arsenic groundwater in the plain, leading to 18% more areas and 33.8% more residents being potentially at risk. This study suggests that the threat of groundwater overexploitation may be much more severe than previously expected.

Occurrence, predictors and hazards of elevated groundwater arsenic across India through field observations and regional-scale AI-based modeling

Existence of wide spread elevated concentrations of groundwater arsenic (As) across South Asia, including India, has endangered a huge groundwater-based drinking water dependent population. Here, using high-spatial resolution As field-observations (~3 million groundwater sources) across India, we have delineated the regional-scale occurrence of elevated groundwater As (≥10 μg/L), along with the possible geologic-geomorphologic-hydrologic and human-sourced predictors that influence the spatial distribution of the contaminant. Using statistical and machine learning method, we also modeled the groundwater As concentrations probability at 1 Km resolution, along with probabilistic delineation of high As-hazard zones across India. The observed occurrence of groundwater As was found to be most strongly influenced by geology-tectonics, groundwater-fed irrigated area (%) and elevation. Pervasive As contamination is observed in major parts of the Himalayan mega-river Indus-Ganges-Brahmaputra basins, however it also occurs in several more-localized pockets, mostly related to ancient tectonic zones, igneous provinces, aquifers in modern delta and chalcophile mineralized regions. The model results suggest As-hazard potential in yet-undetected areas. Our model performed well in predicting groundwater arsenic, with accuracy: 82% and 84%; area under the curve (AUC): 0.89 and 0.88 for test data and validation datasets. An estimated ~90 million people across India are found to be exposed to high groundwater As from field-observed data, with the five states with highest hazard are West Bengal (28 million), Bihar (21 million), Uttar Pradesh (15 million), Assam (8.6 million) and Punjab (6 million). However it can be much more if the modeled hazard is considered (>250 million). Thus, our study provides a detailed, quantitative assessment of high groundwater As across India, with delineation of possible intrinsic influences and exogenous forcings. The predictive model is helpful in predicting As-hazard zones in the areas with limited measurements.

Modeling regional-scale groundwater arsenic hazard in the transboundary Ganges River Delta, India and Bangladesh: Infusing physically-based model with machine learning

For the last few decades, toxic levels of arsenic (As) in groundwater from the aquifers of the Ganges River delta, India and Bangladesh, have been known to cause serious public health concerns. Innumerable studies have advocated the control of geomorphologic, geologic, hydrogeologic, biogeochemical, and anthropogenic factors on arsenic mobilization, flow, and distribution patterns within the Ganges River delta. We have developed transboundary regional-scale models for computing the probability of groundwater As concentrations to exceed the WHO permissible thresholds for drinking water of 10 μg/L within the Ganges River delta as a function of the various geomorphologic-(hydro)geologic-hydrostratigraphic-anthropogenic controlling factors, using statistical methods and artificial intelligence (AI) [i.e., machine learning] techniques namely, Random Forest (RF), Boosted Regression Trees (BRT) and Logistic Regression (LR) algorithms, followed by probabilistic delineation the high As-hazard zones within the delta. A "hybrid multi-modeling approach" was adapted for this study, which involved the introduction of hydrostratigraphic parameters (aquifer connectivity and surficial aquitard thickness) derived from a high-resolution transboundary hydrostratigraphic model developed for the Ganges River delta aquifer system, as predictors for modeling groundwater As probabilities within the delta. The RF model outperforms the BRT and LR model in terms of model performance. Model outputs suggest the dominant influence of surficial aquitard thickness and groundwater-fed irrigated area (%) on groundwater As. While, the north-central and southern regions of the Ganges River delta show low As-hazard (<10 μg/L), the western and north-eastern regions demonstrate elevated hazard level (>10 μg/L). An estimated 30.3 million people are found to be exposed to elevated groundwater As within the study area. Thus, our study demonstrates that such hybrid, predictive models are not only helpful in delineating the regional-scale distribution of groundwater As-hazard zones in the areas with limited As data but is also useful in identifying the possible exogenous forcing that may have led to the worst, natural pollution in human history.

Rising arsenic concentrations from dewatering a geothermally influenced aquifer in central Mexico

This study identifies causes of rising arsenic (As) concentrations over 17 years in an inter-montane aquifer system located just north of the Trans-Mexican-Volcanic-Belt in the Mesa central physiographic region that is extensively developed by long-screened production wells. Arsenic concentrations increased by more than 10 µg/L in 14% (3/22) of re-sampled wells. Similarly, in a larger scale analysis wherein As concentrations measured in 137 wells in 2016 were compared to interpolated, baseline concentrations from 246 wells in 1999, As concentrations rose more than 10 µg/L in 30% of wells. Between 1999 and 2016, the percentage of all wells sampled in each basin-wide sampling campaign exceeding the World Health Organization's 10 µg/L drinking water limit increased from 38 to 64%. Principal Components Analysis (PCA), step-wise multiple regression, and Random Forest modeling (RF) revealed that high As concentrations are closely associated with high pH and temperature, and high concentrations of fluoride (F), molybdenum (Mo), lithium (Li), sodium (Na) and silica (Si), but low calcium (Ca) and nitrate (NO₃) concentrations. Pumping-induced mixing with hot, geothermally impacted groundwater generates alkaline water through hydrolysis of silicate minerals. The rising pH converts oxyanion sorption sites from positive to negative releasing As (and Mo) to pore waters. The negative correlation between nitrate and As concentrations can be explained by conservative mixing of shallow, young groundwater with geothermally influenced groundwater. Therefore water carrying an anthropogenic contaminant dilutes water carrying geogenic contaminants. This process is enabled by long well screens. Over-exploitation of aquifers in geothermal regions for agriculture can drive As concentrations in water from production wells to toxic levels even as the total dissolved solids remain low.

Origin of Groundwater Arsenic in a Rural Pleistocene Aquifer in Bangladesh Depressurized by Distal Municipal Pumping

Across South Asia, millions of villagers have reduced their exposure to high-arsenic (As) groundwater by switching to low-As wells. Isotopic tracers and flow modeling are used in this study to understand the groundwater flow system of a semi-confined aquifer of Pleistocene (>10 kyr) age in Bangladesh that is generally low in As but has been perturbed by massive pumping at a distance of about 25 km for the municipal water supply of Dhaka. A 10- to 15-m-thick clay aquitard caps much of the intermediate aquifer (>40- to 90-m depth) in the 3-km² study area, with some interruptions by younger channel sand deposits indicative of river scouring. Hydraulic heads in the intermediate aquifer below the clay-capped areas are 1-2 m lower than in the high-As shallow aquifer above the clay layer. In contrast, similar heads in the shallow and intermediate aquifer are observed where the clay layer is missing. The head distribution suggests a pattern of downward flow through interruptions in the aquitard and lateral advection from the sandy areas to the confined portion of the aquifer. The interpreted flow system is consistent with ³H-³He ages, stable isotope data, and groundwater flow modeling. Lateral flow could explain an association of elevated As with high methane concentrations within layers of gray sand below certain clay-capped portions of the Pleistocene aquifer. An influx of dissolved organic carbon from the clay layer itself leading to a reduction of initially orange sands has also likely contributed to the rise of As.

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.

A meta-analysis of the distribution, sources and health risks of arsenic-contaminated groundwater in Pakistan

Globally, millions of people who rely on groundwater for potable purposes and agriculture have been inadvertently exposed to toxic arsenic (As) because of its natural occurrence in groundwater in several countries of Asia, Europe and America. While the presence of As in groundwater and its impacts on human health have been documented in many countries, there is little information on As contamination in Pakistan. This review highlights, for the first time, the extent and severity of As-induced problems in Pakistan based on relevant published papers; discusses possible sources of As contamination of aquifers; and estimates As-induced potential health hazards in the country in relation to global data. Data from 43 studies (>9882 groundwater samples) were used to describe As variability in groundwater of Pakistan and for comparison with global data. The mean groundwater As content reported in these studies was 120 μg/L (range: 0.1-2090 μg/L; SD: ±307). About 73% of the values for mean As contents in the 43 studies were higher than the World Health Organization (WHO) permissible limit (10 μg/L) for drinking water, while 41% were higher than the permissible limit of As in Pakistan (50 μg/L). It was observed that groundwater samples in some areas of Punjab and Sindh provinces contained high As concentrations which were almost equal to concentrations reported in the most contaminated areas of the world. We predicted that the mean values of ADD, HQ and CR were 4.4 μg kg^-1day^-1 (range: 0-77 μg kg^-1day^-1), 14.7 (range: 0-256) and 0.0029 (range: 0-0.0512), respectively, based on mean As concentrations reported in Pakistan. In addition, this article proposes some integrated sustainable solutions and future perspectives keeping in view the regional and global context, as well as the on-ground reality of the population drinking As-contaminated water, planning issues, awareness among civil society and role of the government bodies. Based on available data, it is predicted that almost 47 million people in Pakistan are residing in areas where more than 50% of groundwater wells contain As concentrations above the WHO recommended limit of As in drinking water.

Groundwater quality in the alluvial aquifer system of northwest India: New evidence of the extent of anthropogenic and geogenic contamination

Groundwater depletion has been widely studied in northwest India, but water quality concerns are still poorly constrained. In this study, we explore the hydrochemistry of the top 160m of the aquifer system, through detailed field studies in the Bist-Doab region, considering both anthropogenic and geogenic controls. A detailed comparison is made between sites dominated by urban and agricultural landuse. Salinity, nitrate, chloride and lead concentrations are significantly higher in the shallow (0-50m) groundwater system due to surface anthropogenic contaminant loading from agricultural and urban sources. The widespread occurrence of oxic groundwater within the aquifer system means that denitrification potential is limited and also enhances the mobility of selenium and uranium in groundwater. Geogenic trace elements (e.g. As, Se, F), are generally found at concentrations below WHO guideline drinking water values, however elevated U concentrations (50-70μg/L) are found within the deeper part of the aquifer and shallow urban aquifers associated with higher bicarbonate waters. Higher concentration of Se (10-40μg/L) are found exclusively in the shallow groundwater system where Se is mobilised from soils and transported to depth in the shallow aquifer due to the prevailing oxidising aquifer conditions. New evidence from a range of environmental tracers shows elevated concentrations of anthropogenic contaminants in the deeper part of the aquifer (50-160m deep) and demonstrates vulnerability to vertical migration of contaminants. Continued intensive groundwater abstraction from >100m deep means that water quality risks to the deep aquifer system need to be considered together with water quantity constraints.

Groundwater arsenic contamination in Burkina Faso, West Africa: Predicting and verifying regions at risk

Arsenic contamination in groundwater from crystalline basement rocks in West Africa has only been documented in isolated areas and presents a serious health threat in a region already facing multiple challenges related to water quality and scarcity. We present a comprehensive dataset of arsenic concentrations from drinking water wells in rural Burkina Faso (n=1498), of which 14.6% are above 10μg/L. Included in this dataset are 269 new samples from regions where no published water quality data existed. We used multivariate logistic regression with arsenic measurements as calibration data and maps of geology and mineral deposits as independent predictor variables to create arsenic prediction models at concentration thresholds of 5, 10 and 50μg/L. These hazard maps delineate areas vulnerable to groundwater arsenic contamination in Burkina Faso. Bedrock composed of schists and volcanic rocks of the Birimian formation, potentially harbouring arsenic-containing sulphide minerals, has the highest probability of yielding groundwater arsenic concentrations >10μg/L. Combined with population density estimates, the arsenic prediction models indicate that ~560,000 people are potentially exposed to arsenic-contaminated groundwater in Burkina Faso. The same arsenic-bearing geological formations that are positive predictors for elevated arsenic concentrations in Burkina Faso also exist in neighbouring countries such as Mali, Ghana and Ivory Coast. This study's results are thus of transboundary relevance and can act as a trigger for targeted water quality surveys and mitigation efforts.

Hydrogeological typologies of the Indo-Gangetic basin alluvial aquifer, South Asia

The Indo-Gangetic aquifer is one of the world's most important transboundary water resources, and the most heavily exploited aquifer in the world. To better understand the aquifer system, typologies have been characterized for the aquifer, which integrate existing datasets across the Indo-Gangetic catchment basin at a transboundary scale for the first time, and provide an alternative conceptualization of this aquifer system. Traditionally considered and mapped as a single homogenous aquifer of comparable aquifer properties and groundwater resource at a transboundary scale, the typologies illuminate significant spatial differences in recharge, permeability, storage, and groundwater chemistry across the aquifer system at this transboundary scale. These changes are shown to be systematic, concurrent with large-scale changes in sedimentology of the Pleistocene and Holocene alluvial aquifer, climate, and recent irrigation practices. Seven typologies of the aquifer are presented, each having a distinct set of challenges and opportunities for groundwater development and a different resilience to abstraction and climate change. The seven typologies are: (1) the piedmont margin, (2) the Upper Indus and Upper-Mid Ganges, (3) the Lower Ganges and Mid Brahmaputra, (4) the fluvially influenced deltaic area of the Bengal Basin, (5) the Middle Indus and Upper Ganges, (6) the Lower Indus, and (7) the marine-influenced deltaic areas.

Recharge of low-arsenic aquifers tapped by community wells in Araihazar, Bangladesh, inferred from environmental isotopes

More than 100,000 community wells have been installed in the 150-300 m depth range throughout Bangladesh over the past decade to provide low-arsenic drinking water (<10 μg/L As), but little is known about how aquifers tapped by these wells are recharged. Within a 25 km² area of Bangladesh east of Dhaka, groundwater from 65 low-As wells in the 35-240 m depth range was sampled for tritium (³H), oxygen and hydrogen isotopes of water (¹⁸O/¹⁶O and ²H/¹H), carbon isotope ratios in dissolved inorganic carbon (DIC, ¹⁴C/¹²C and ¹³C/¹²C), noble gases, and a suite of dissolved constituents, including major cations, anions, and trace elements. At shallow depths (<90 m), 24 out of 42 wells contain detectable ³H of up to 6 TU, indicating the presence of groundwater recharged within 60 years. Radiocarbon (¹⁴C) ages in DIC range from modern to 10 kyr. In the 90-240 m depth range, however, only 5 wells shallower than 150 m contain detectable ³H (<0.3 TU) and ¹⁴C ages of DIC cluster around 10 kyr. The radiogenic helium (⁴He) content in groundwater increases linearly across the entire range of ¹⁴C ages at a rate of 2.5×10^-12 ccSTP ⁴He g^-1 yr^-1. Within the samples from depths >90 m, systematic relationships between ¹⁸O/¹⁶O, ²H/¹H, ¹³C/¹²C and ¹⁴C/¹²C, and variations in noble gas temperatures, suggest that changes in monsoon intensity and vegetation cover occurred at the onset of the Holocene, when the sampled water was recharged. Thus, the deeper low-As aquifers remain relatively isolated from the shallow, high-As aquifer.