Advertisement

Hydrogeology Journal

, Volume 26, Issue 5, pp 1573–1589 | Cite as

Redox zonation for different groundwater flow paths during bank filtration: a case study at Liao River, Shenyang, northeastern China

  • Xiaosi Su
  • Shuai Lu
  • Wenzhen Yuan
  • Nam Chil Woo
  • Zhenxue Dai
  • Weihong Dong
  • Shanghai Du
  • Xinyue Zhang
Paper
  • 330 Downloads

Abstract

The spatial and temporal distribution of redox zones in an aquifer is important when designing groundwater supply systems. Redox zonation can have direct or indirect control of the biological and chemical reactions and mobility of pollutants. In this study, redox conditions are characterized by interpreting the hydrogeological conditions and water chemistry in groundwater during bank infiltration at a site in Shenyang, northeast China. The relevant redox processes and zonal differences in a shallow flow path and deeper flow path at the field scale were revealed by monitoring the redox parameters and chemistry of groundwater near the Liao River. The results show obvious horizontal and vertical components of redox zones during bank filtration. Variations in the horizontal extent of the redox zone were controlled by the different permeabilities of the riverbed sediments and aquifer with depth. Horizontally, the redox zone was situated within 17 m of the riverbank for the shallow flow path and within 200 m for the deep flow path. The vertical extent of the redox zone was affected by precipitation and seasonal river floods and extended to 10 m below the surface. During bank filtration, iron and manganese oxides or hydroxides were reductively dissolved, and arsenic that was adsorbed onto the medium surface or coprecipitated is released into the groundwater. This leads to increased arsenic content in groundwater, which poses a serious threat to water supply security.

Keywords

Bank filtration Redox zonation China Hydrochemistry Arsenic 

Zonation d’oxydoréduction pour différentes voies d’écoulement d’eaux souterraines lors du processus de filtration par Berge: une étude de cas Sur la rivière Liao, Shenyang, Nord-Est de la Chine

Résumé

La distribution spatiale et temporelle des zones d’oxydoréduction (redox) dans un aquifère est. importante lors de la conception de systèmes d’alimentation en eau souterraine. La zonation redox peut contrôler directement ou indirectement les réactions biologiques et chimiques et la mobilité des contaminants. Dans cette étude, les conditions redox sont caractérisées par l’interprétation des conditions hydrogéologiques et la chimie de l’eau des eaux souterraines au cours de l’infiltration par berge sur le site de Shenyang, Nord Est de la Chine. Les processus redox concernés et les différences zonales dans une voie d’écoulement peu profond et une plus profonde à l’échelle du site ont été mis en évidence grâce au suivi des paramètres redox et la chimie des eaux souterraines près de la rivière Liao. Les résultats montrent d’évidentes composantes horizontales et verticales des zones d’oxydoréduction au cours de la filtration sur berge. Les variations horizontales de la zone d’oxydoréduction sont contrôlées par les différentes conductivités hydrauliques des sédiments du lit de la rivière et de l’aquifère en profondeur. Dans un plan horizontal, la zone redox se trouvait à moins de 17 m de la berge pour la voie d’écoulement peu profonde et à moins de 200 m pour la voie d’écoulement profonde. Dans le plan vertical, la zone redox est. influencée par les précipitations et les inondations saisonnières de la rivière et s’étend à 10 m sous la surface. Au cours de l’infiltration par berge, les oxydes ou hydroxydes de fer et de manganèse sont dissous par réduction, et l’arsenic qui a été adsorbé à la surface du milieu ou coprécipité est. relargué dans les eaux souterraines. Cela conduit à une augmentation de la teneur en arsenic dans les eaux souterraines, ce qui constitue une sérieuse menace pour la sécurité de l’approvisionnement en eau potable.

Zonificación redox para diferentes trayectorias de flujo de aguas subterráneas durante la filtración de banco: un estudio de Caso en el río Liao, Shenyang, noreste de China

Resumen

La distribución espacial y temporal de las zonas redox en un acuífero es importante cuando se diseñan sistemas de suministro de agua subterránea. La zonación redox puede tener un control directo o indirecto de las reacciones biológicas y químicas y la movilidad de los contaminantes. En este estudio, las condiciones redox se caracterizan por interpretar las condiciones hidrogeológicas y la química del agua en las aguas subterráneas durante la infiltración del banco en un sitio en Shenyang, al noreste de China. Los procesos redox relevantes y las diferencias zonales en una trayectoria de flujo poco profundo y una trayectoria de flujo más profundo en la escala de campo se revelaron mediante el monitoreo de los parámetros redox y la química del agua subterránea cerca del río Liao. Los resultados muestran componentes horizontales y verticales obvios de las zonas redox durante la filtración de banco. Las variaciones en la extensión horizontal de la zona redox fueron controladas por las diferentes permeabilidades de los sedimentos del lecho del río y del acuífero según la profundidad. Horizontalmente, la zona redox estaba situada a 17 m de la margen del río para la trayectoria de flujo superficial y dentro de los 200 m para la trayectoria de flujo profundo. La extensión vertical de la zona redox se vio afectada por la precipitación y las inundaciones estacionales del río y se extendió a 10 m por debajo de la superficie. Durante la filtración de banco, los óxidos o hidróxidos de hierro y manganeso se disolvieron en forma reductiva, y el arsénico que se adsorbió sobre la superficie del medio o se coprecipitó se libera al agua subterránea. Esto conduce a un mayor contenido de arsénico en las aguas subterráneas, lo que representa una grave amenaza para la seguridad del suministro de agua.

河岸入渗期间不同地下水流通道的氧化还原反应成带现象:中国东北沈阳辽河的一个研究案例

摘要

当设计地下水供水系统时,含水层中氧化还原反应带的空间和时间分布非常重要。氧化还原反应成带现象可直接或间接控制生物和化学反应以及污染物的迁移。在本研究中,通过解译中国东北沈阳某地河岸入渗期间地下水中的水文地质条件和水化学过程,描述了氧化还原反应的特征。通过监测辽河附近地下水中氧化还原反应的参数及化学过程揭示了野外尺度的浅层水流通道及深部水流通道相关的氧化还原反应过程和分区差异。结果显示出,河岸入渗期间有明显的水平和垂直痕迹。氧化还原反应带的水平范围上的变化随深度变化受到河床沉积层和含水层不同渗透率的控制。水平上,对于浅层水流通道,氧化还原反应带位于河岸17 米之内,而对于深层水流通道,则其位于河岸200米以内。氧化环氧反应带垂直范围受到降水和季节性河流洪水的影响,可延伸到地表以下10米。在河岸入渗期间,铁锰氧化物和或氢氧化物还原溶解,吸附在介质表面或者沉淀的砷释放到地下水中。这导致地下水中的砷含量升高,对供水安全产生严重的威胁。

Zona redox em diferentes caminhos do fluxo de água subterrânea durante a filtração em margem: um estudo de Caso no rio Liao, Shenyang, nordeste da China

Resumo

A distribuição especial e temporal das zonas redox em um aquífero é importante quando se projetam os sistemas de abastecimento por águas subterrâneas. Zonas redox podem controlar direta ou indiretamente as reações químicas e biológicas e a mobilidade dos poluentes. Neste estudo, as condições redox são caracterizadas pela interpretação das condições hidrogeológicas e químicas das águas subterrâneas durante a infiltração pelas margens em um local da cidade de Shenyang, nordeste da China. Os relevantes processos redox e diferentes zonas no percurso de fluxo, sejam rasos ou profundos, na escala de campo, foram revelados por monitoramento de parâmetros redox e química das águas subterrâneas próximo ao Rio Liao. Os resultados mostram claramente componentes verticais e horizontais das zonas redox ao longo da filtração em margem. Variações na extensão horizontal da zona redox foram controladas por diferentes permeabilidades dos sedimentos do leito do rio e do aquífero em profundidade. Horizontalmente, a zona redox situa-se a 17 m das margens do rio no percurso de fluxo raso e cerca de 200 m no percurso de fluxo profundo. A extensão vertical da zona redox foi afetada pela precipitação e inundações sazonais do rio e alcançou até 10 m abaixo da superfície. Durante a filtração em margem, óxidos e hidróxidos de ferro e manganês foram dissolvidos e reduzidos, e o arsênio, que foi adsorvido pela superfície do meio ou coprecipitado, foi liberado para as águas subterrâneas. Isto levou ao aumento de concentrações de arsênio nas águas subterrâneas, o que representa sério risco a segurança do abastecimento de água.

Introduction

Bank filtration induced by groundwater pumping adjacent to a perennial river is considered to be an efficient and natural treatment technology for water quality improvement (Bouwer 2002; Hamann et al. 2016; Hu et al. 2016; Kedziorek and Bourg 2009; Polomčić et al. 2013). Bank filtration facilitates the protection of water quality because it can attenuate or degrade pollutants. These pollutants can include suspended solids, inorganic or organic substances, poisonous heavy metals, pathogenic viruses, parasites, and bacteria. Physical, chemical, and biological filtration occurs when the river water passes through the riverbed sediments and aquifer media. Bank filtration has been successfully applied for more than 140 years worldwide (Gandy et al. 2007; Hiscock and Grischek 2002; Huntscha et al. 2013; Massmann et al. 2006; Tufenkji et al. 2002; Worch et al. 2002).

Riverbed sediments provide the first zone for river- filtration induced by groundwater pumping; however, the infiltrated water has different flow paths in the aquifer because of the lithological and structural controls of the riverbed sediments. Different flow paths result in different seepage times through the filtration zone. Flow paths also affect water–rock reactions and biogeochemical processes, change the degradation rates of biodegradable compounds, and cause water quantity and quality differences at different depths of the aquifer, resulting in the stratification of water quality (Hancock 2002; Haque and Johannesson 2006; Massmann et al. 2008a; Xie et al. 2013). For instance, redox reactions during bank filtration may cause metals and metalloids, such as iron, manganese, and arsenic—which have significant redox sensitivity (Appelo and Postma 2005; Farnsworth and Hering 2011)—to occur in infiltrating river water. Changes in hydrodynamic conditions lead to a differentiation in the distinguishable redox zones. This zonation tends to occur in the order of decreasing free-energy yield of reactions involving electron acceptors (such as oxygen, nitrate, iron and manganese oxides or hydroxides, and sulfate). Redox zonation during bank filtration will be at different scales and have different zonation characteristics (Jung et al. 2015; Massmann et al. 2004, 2008b; Scow and Hicks 2005; Stuyfzand 2011; von Rohr et al. 2014). The redox processes can alternately mobilize or immobilize potentially toxic metals associated with minerals occurring in riverbed sediments and aquifer media; thus, it is important to investigate redox zonation during bank filtration to improve understanding of the migration of river pollutants and their removal, as well as the release of poisonous metals in aquifers to groundwater (Hancock 2002; Kedziorek et al. 2008).

Previous studies of redox zonation during bank filtration have investigated how oxygenated river water infiltrates into anoxic aquifers (Brown et al. 2000; Lovley and Phillips 1988; Massmann et al. 2008b) and most studies have focused on redox processes and the redox state of the infiltrate within a few meters (Bertelkamp et al. 2016; Heberer et al. 2008; Lewandowski et al. 2011; Prommer and Stuyfzand 2005; Sharma et al. 2012). However, until now there have been few detailed field-scale investigations of redox zonation in different flow paths during bank filtration controlled by different hydrogeological conditions (Bourg and Bertin 1993; Massmann et al. 2008b); therefore, this study addressed this knowledge gap by monitoring and investigating the redox zonation in different flow paths during bank filtration at the field scale. The results will be of great value for management plans to ensure the sustainability of drinking-water-production systems.

The study area is located on the south bank of the Liao River, northeastern China. The aquifer medium of the water source area is rich in iron and manganese primary minerals, whereby the iron and manganese content in groundwater generally exceeds maximum allowable levels (China 2006). Iron and manganese in groundwater are very sensitive to redox conditions during bank filtration and this sensitivity greatly facilitates reductive dissolution in the nearshore zone and oxidative precipitation near the pumping well. This sensitivity also facilitates the release of toxic arsenic to groundwater, causing water pollution and the blockage of pumping wells, resulting in reduced exploitation efficiency. In this study, redox and other chemical parameters associated with groundwater proximate to the Liao River were monitored to characterize the redox conditions during bank infiltration. Thus, the aim of the study was to reveal the relevant redox processes and the zonal differences in different flow paths at the field scale.

Materials and methods

Site description

The site is situated 40 km north of the city of Shenyang, northeastern China (Fig. 1). It is located on the alluvial and diluvial plain of the Liao River and is characterized by flat topography, with land-surface elevations varying from 42.0 to 52.0 m above sea level. The area has a typical temperate, semi-humid monsoon climate and the mean annual temperature is 7 °C, with a maximum temperature of 27 °C (July) and a minimum temperature of −19 °C (January). The mean annual precipitation is 635.5 mm and more than 80% of the precipitation is concentrated in the flood season (May to September), while the mean annual surface-water evaporation is 1,594 mm. The Liao River, which flows through the northern part of the study area, has a width of 100.0–300.0 m and a runoff of 7.9–41.8 m3/s (Zhang 2015).
Fig. 1

Location of the study area

Groundwater is stored in an unconsolidated phreatic Quaternary aquifer, which overlies impermeable Tertiary glutenite. Overlying the aquifer is loam and mucky soil with a thickness of 1.0–3.0 m (Fig. 2). The aquifer is about 50 m thick and is composed of fine sand, medium-coarse sand containing gravel, sandy gravel, pebbly gravel, and coarse sand containing gravel. There are 12 pumping wells in the study area, and the total amount of groundwater pumped is approximately 30,000 m3/day, which has remained stable for many years. As a result, a groundwater depression cone has formed with a depth of 3.0 m in the center. The surrounding groundwater flows toward this center. The groundwater is always recharged by Liao River water.
Fig. 2

Hydrogeological cross-section and groundwater flow paths

Near the south bank of the Liao River, there is a relatively dense, relatively continuous muddy layer, with a thickness of about 0–0.7 m. In the middle of the riverbed, the horizontal continuity and stratification of the sediment is low, and the interlayer space is relatively large, indicating a relatively high permeability. In the periphery of the groundwater depression cone area, the hydraulic gradient between the river and aquifer acts as a driving force. Consequently, gravity flow to the center of the depression cone induces groundwater flow from the river to the pumping wells. There are two different primary flow paths from the river to the center of the depression cone (Fig. 2). The first involves shallow, lateral, low-permeability infiltration. In this case, the hydraulic conductivity is about 13.5 m/day, which inhibits the speed of infiltration, generates longer migration times, and increases chemical reactions. The second flow path involves deep, vertical, high-permeability infiltration. The hydraulic conductivity is about 63.9 m/day, which creates longer and larger migration paths and a larger specific surface area in the reaction process (Su et al. 2017).

The aquifer medium at 5.0–45.0 m below the ground surface is mainly composed of feldspar and clay minerals. The combined contents of quartz, potash feldspar, and plagioclase form 64–94% of the aquifer medium, illite-smectite layer-mixed minerals 4–36%, and kaolinite and mica about 2%. Typical iron content is 2,809–4,261 and 2,532–3,170 mg/kg in the shallow and deep groundwater, respectively. The manganese content is 1,401–1,805 and 1,600–1,665 mg/kg in the shallow and deep groundwater, respectively.

Sampling

The monitoring section was orientated approximately along the groundwater flow direction between the south bank of the Liao River and pumping wells Nos. 5 and 6 (the center of the regional depression cone; Fig. 3). Monitoring sites HJ1, HJ2, HJ3, and HJ4 were established at distances of 17.0, 200.0, 350.0, and 560.0 m from the south bank of the river, respectively, whreeby these four sites were each screened to 10.0, 20.0, 35.0, and 45.0 m below the ground surface. A high-resolution stratified dynamic monitoring network was established at the nearshore zone because this area has the largest changes in biogeochemical environment, and is the most sensitive to pH and redox. Monitoring sites HB1, HB2, HB3, and HB4, screened to depths of 6.0, 7.0, 8.0, 9.0, 10.0, and 11.0 m below ground surface at each site, were established along the groundwater flow direction from the riverbank to HJ1 at distances of 1.5, 4.0, 6.5, and 12.5 m from the south bank of the river, respectively. Monitoring site HC, screened to depths of 4.0, 5.0, 7.0, and 9.0 m, and referred to as HC5, HC6, HC7, and HC8, was established in the river at a distance of 2.0 m north from the south bank of the river. The main well monitoring parameters are shown in Table 1. Monitoring wells of different depths were established at each monitoring site so that the hydrochemical and other relevant information could be obtained for depths. Water samples were collected for hydrochemistry analysis from the river and all the monitoring wells monthly from May 2014 to December 2015, except for November 2014 and March 2015. The sediment samples were collected in December 2015 for soil organic carbon (SOC) analysis.
Fig. 3

Distribution of monitoring stations and isolines of groundwater levels

Table 1

Well depths and construction details

Well No.

Hole depth (m)

Depth of screen sections (m)

Seal depths (m)

Bottom seal

Top seal

HJ1–1\HJ2–1\HJ3–1\HJ4–1

10.0

9.0–4.0

10.0–9.0

4.0–3.0

HJ1–2\HJ2–2\HJ3–2\HJ4–2

20.0

19.0–14.0

20.0–19.0

14.0–13.0

HJ1–3\HJ2–3\HJ3–3\HJ4–3

35.0

34.0–29.0

35.0–34.0

29.0–28.0

HJ1–4\HJ2–4\HJ3–4\HJ4–4

45.0

44.0–39.0

45.0–44.0

39.0–38.0

HB1–2\HB2–2\HB3–2\HB4–2

6.0

5.0–4.0

6.0–5.0

4.0–3.0

HB1–3\HB2–3\HB3–3\HB4–3

7.0

6.0–5.0

7.0–6.0

5.0–4.0

HB1–4\HB2–4\HB3–4\HB4–4

8.0

7.0–6.0

8.0–7.0

6.0–5.0

HB1–5\HB2–5\HB3–5\HB4–5

9.0

8.0–7.0

9.0–8.0

7.0–6.0

HB1–6\HB2–6\HB3–6\HB4–6

10.0

9.0–8.0

10.0–9.0

8.0–7.0

HB1–7\HB2–7\HB3–7\HB4–7

11.0

10.0–9.0

11.0–10.0

9.0–8.0

HC5

4.0

3.0–2.0

4.0–3.0

2.0–1.0

HC6

5.0

4.0–3.0

5.0–4.0

3.0–2.0

HC7

7.0

6.0–4.0

7.0–6.0

4.0–3.0

HC8

9.0

8.0–6.0

9.0–8.0

6.0–5.0

Monitoring wells were purged for at least 3–5 well volumes before sampling. Once the temperature, pH, electrical conductivity (EC), oxidation-reduction potential (Eh), and dissolved oxygen (DO) values stabilized, the groundwater samples were collected using a Proactive Mega-Typhoon pump (Proactive Environmental Products International LLC, Florida, USA). The pump was equipped with a controller to achieve stable low-flow sampling (minimum sampling speed of 0.95 L/min). This prevents changes in water oxidation-reduction-sensitive indicators, such as Eh and DO, during sampling. River-water samples were sampled by syringe directly from the river at a mean depth of 20.0 cm below the water surface. Water samples were filtered onsite using 0.45-μm membrane filters. Prior to sample collection, all bottles used were washed thoroughly. Samples for anion-cation ions and arsenic analysis were collected in 120.0-ml pre-cleaned polyethylene bottles. Samples for anion-cation ions analysis were not acidified, while samples for arsenic analysis were immediately acidified with HCl to pH < 2 and sealed without headspace. Samples for dissolved organic carbon (DOC) analysis were collected in 120.0-ml pre-cleaned amber glass bottles, immediately acidified with H2SO4 to pH < 2, and sealed without headspace. Checks were done to ensure that there were no bubbles or headspace present in the bottles upon sample collection. Duplicate samples were collected for 30% of the samples.

Undisturbed soil samples were collected from monitoring points (Table 2) by a Beeker portable sediment sampler (Eijkelkamp Agrisearch Equipment, Giesbeek, Netherlands) and a percussion drill. The samples were stored in 120.0-ml amber glass bottles without headspace.
Table 2

The distribution of soil organic carbon (SOC)

Location

Distance to the riverbank (m)

Depth (m)

SOC content (g/kg)

HS2

−6.0

2.0

15.83

HC

−2.0

4.0

14.88

HB1

1.5

5.0

5.23

HB2

4.0

6.0

3.37

HB3

6.5

7.0

3.18

HB4

12.5

8.0

2.57

HJ1a

17.0

10.0

2.49

HJ1b

17.0

35.0

2.41

HJ2

200.0

35.0

1.32

HJ3

350.0

35.0

1.05

HJ4

560.0

35.0

1.01

Test methods

Parameters such as water level, temperature, pH, EC, Eh, DO, and TDS were determined at the wellheads using a Hach HQ40d portable tester (Hach Company, Colorado, USA), and each of these parameters was recorded until the reading had stabilized. Iron, manganese, and sulfide were measured onsite using a Hach DR2800 spectrophotometer; at least three measurements were taken for each of these parameters and then an average was obtained. Anions and cations, such as K+, Na+, Ca2+, Mg2+, Cl, SO42−, and NO3, were analyzed by ion chromatography (881 Compact IC, Metrohm, Herisau, Switzerland). The CO32− and HCO3 contents were measured using acid-base titration. Arsenic concentration analysis was conducted using a hydride generation atomic fluorescence spectrophotometer (HG-AFS; AFS-820, Beijing Titan Instruments Co., Ltd., Beijing, China). DOC and SOC were analyzed using a total organic carbon analyzer (TOC-L CPH CN200, Shimadzu, Kyoto, Japan). These tests, which included blank controls, were conducted within one week.

Results and discussion

The Liao River water was in a state of oxidation (Eh was about 100.0 mV on average, and the concentration of DO was about 13.0 mg/L on average). The electron donor (DOC) levels were more than 31.03 mg/L-C on average, clearly outweighing the electron acceptor (O2, NO3, SO42−) levels, while Fe2+/Mn2+ concentrations in river water were below the detection limits (Table 3); however, the levels of electron acceptors such as total Fe (4.60–25.20 mg/L) and total Mn (0.92–6.58 mg/L) in groundwater exceeded the World Health Organization (WHO) standards (≤ 0.3 and ≤0.1 mg/L for total Fe and total Mn, respectively). There were also high levels of As (0–25.0 × 10−3 mg/L) in groundwater, indicating that water-quality safety standards had been exceeded and that there was a threat to human health (Table 4; China 2006; European Commission 1998; USEPA 2011; WHO 2011).
Table 3

Redox relevant species in the river

Units

Electron acceptors

Electron donor

O2

NO3

SO42−

Sum

DOC

mg/L

9.01

6.99

65.49

31.03

mmol/L

0.28

0.11

0.68

1.03

Meq/L

1.13

0.34

1.36

2.83

4.14

Table 4

Excessive components in groundwater

Items for comparison

pH

Total hardness (CaCO3) (mg/L)

TDS (mg/L)

Total Fe (mg/L)

Total Mn (mg/L)

CODMn (mg/L)

As (mg/L)

China Standards for Drinking Water Quality (GB5749–2006)

6.5–8.5

≤ 450

≤ 1000

≤ 0.3

≤ 0.1

≤ 3

≤ 0.01

World Health Organization standards

6.5–8.5

≤ 500

≤ 1000

≤ 0.3

≤ 0.1

≤ 0.01

European Union standards

6.5–9.5

≤ 0.2

≤ 0.05

≤ 0.01

United States Environmental Protection Agency standards

6.5–8.5

≤ 500

≤ 0.3

≤ 0.05

≤ 0.01

Contents in the study area

6.49–8.40

116.09–285.76

217.5–746.0

4.60–25.20

0.92–6.58

0–17.18

0–25.0 × 10−3

Exceeding standard rate in the study area (%)

100

100

93.75

12.5

The chemical constituents sensitive to redox changes are widely involved in the degradation of organic carbon with electron acceptors. Degradation proceeds in a sequential order based on the energy, and to some extent, the redox parameters that influence the spatial distribution of redox zones (Diem et al. 2013; Greskowiak et al. 2006; Henzler et al. 2016). The sequential order is oxygen (O2), nitrate (NO3), manganese (Mn(IV)) and iron (Fe(III)) oxides or hydroxides, sulfate (SO42−), and then carbon dioxide (CO2). Based on a long sequence of dynamic monitoring data, the characteristics of redox parameter changes in the river water and groundwater were analyzed.

Variations in the content of some chemical constituents sensitive to redox changes along the shallow flow path

Eh

Eh in the river was high (about 0–190.0 mV), indicating that the river formed an oxidizing environment, but it decreased gradually along the flow path (Fig. 4). At HJ1-1 (17.0 m away from the riverbank), the Eh in groundwater had decreased and was −190.0 to −110.0 mV. This was a typical reducing environment, reflecting the change in redox conditions from an oxidizing to a reducing environment. At HJ3-1 (350.0 m away from the riverbank and near the pumping wells), the Eh level elevated slightly, indicating that oxygen had entered the groundwater system due to air traps. It is inferred that the increase in Eh value near the pumping wells was the result of fluctuations in groundwater levels caused by the operation of the pumping wells. The changes in acid-base and redox conditions in the groundwater were the most obvious in the shallow flow path from the riverbank to 17.0 m away from the riverbank.
Fig. 4

Boxplots of redox zonation indicators for the shallow flow path: a Eh, b DOC, c DO, d NO3, e Mn2+, f Fe2+, g SO42−

DOC

The concentration of DOC in the river was high (about 7.0–38.0 mg/L), but after the river water infiltrated the groundwater, the concentration decreased gradually along the flow path away from the river. DOC acts as an electron donor and is often involved in oxidation-reduction reactions. At HJ1-1 (17.0 m away from the riverbank), the concentration of DOC decreased to about 1.0 mg/L, and was below the detection limit at HJ2-1 (200.0 m away from the riverbank). In general, the most distinct changes in DOC occurred within 17.0 m of the riverbank. The seasonal variation in DOC content in groundwater was similar to that of river water, and the content in the winter season (December 2014 to March 2015) was higher than that in the summer season (June to September 2015). As the distance to the riverbank increased, the seasonal variation in DOC content in groundwater decreased, and the content became relatively stable (about 2.0 mg/L; Fig. 5). In the infiltration process, the DOC carried by the river was gradually depleted. In contrast, SOC in the riverbed sediment was as high as 15.83 g/kg and it participated in the reaction as electron donor, providing the driving force for the reaction (Fig. 6; Table 2); therefore, SOC and DOC acted together as electron donors.
Fig. 5

Seasonal variations of dissolved organic carbon (DOC) in river water (HS2) and groundwater (HJ1-1, HJ2-1, HJ3-1, HJ4-1)

Fig. 6

Soil organic carbon (SOC) contents of aquifer sediments

DO

The concentration of DO in the river was high (about 6.5–18.0 mg/L) and reflected oxidizing conditions, but decreased gradually along the groundwater flow path away from the river. As distance from the riverbank further increased, the groundwater formed a reducing environment, as indicated by the continually decreasing DO concentration. At HB2 (4.0 m away from the riverbank), the DO was very low, indicating that the main oxygen consumption zone was within 4.0 m of the riverbank.

NO3

The concentration of NO3 in the river (8.0–23.5 mg/L) was higher than that in the groundwater, and decreased gradually along the flow path to almost below the detection limit at HB3 (6.5 m away from the riverbank).

Mn2+

The concentration of Mn2+ in the river was almost below the detection limit; however, as the distance to the riverbank increased, Mn(IV) reduction occurred, and the concentration of Mn2+ in the groundwater clearly increased. The Mn2+ content peaked at HB4 (12.5 m away from the riverbank), which reflected an active zone containing manganese minerals; although, further along the flow path, the Mn2+ content decreased.

Fe2+

The concentration of Fe2+ in the river was almost below the detection limit. Reductive dissolution of iron-bearing minerals occurred with increasing distance from the riverbank, resulting in increased Fe2+ content in the groundwater (Yuan 2017). The Fe2+ content peaked at HJ1-1 (17.0 m away from the riverbank), which reflected the active zone of iron minerals. Beyond HJ1-1, the Fe2+ content decreased as distance from the riverbank increased.

The geochemical characteristics at HJ1 showed clear changes in the spatial distribution of the main redox parameters; therefore, this monitoring site was used to demonstrate the seasonal variation of Mn2+ and Fe2+ contents in the shallow and deep groundwater during the monitoring period from January to September in 2015. The contents of Mn2+ and Fe2+ in groundwater were lower in summer and higher in winter (Fig. 7). This seasonal variation could have been caused by two main factors—the first being that the DOC content in river water and groundwater also had higher values in winter and lower values in summer during the same period (Fig. 5). Thus, the river water could have provided more electron donors in winter, and consequently more Mn2+ and Fe2+ was produced through the reduction process. Alternatively, the hydrological gradient between the river and groundwater in winter was relatively small. Thus, the flow velocity was slower and the residence time of infiltrated river water in the aquifer was longer, which increased the reduction time for manganese and iron minerals, and consequently increased the content of Mn2+ and Fe2+ in groundwater in winter.
Fig. 7

Seasonal percentage variation of total manganese (TMn), total iron (TFe), Mn2+ and Fe2+ in wells a HJ1–1 and b HJ1–3

SO4 2−

The concentrations of SO42− in the river water were high (about 50.0–88.0 mg/L), and decreased gradually along the flow path away from the river to below the detection limit at HJ1-1 (17.0 m away from the riverbank). This was in contrast to the sulfide change trend, whereby the sulfide content was low in the river, and gradually increased with increased distance from the riverbank and reached the peak (about 11.0 × 10−3 mg/L on average) at HJ1-1 (Yuan 2017).

Variations in the content of some chemical constituents sensitive to redox changes along the deep flow path

DOC

In the deep part of the aquifer, the groundwater flow rate was high, and the nutrient flux was high. The concentration of DOC was clearly reduced at HJ1-3 (17.0 m away from the riverbank) (Fig. 8) and the DO and NO3 contents at HJ1-3 were almost nil. Thus, the main changes in DOC concentration occurred between the riverbank and 17.0 m from the riverbank in the deep groundwater, and this was the main O2/NO3 reduction zone.
Fig. 8

Boxplots of redox zonation indicators for the deep flow path: a DOC, b DO, c NO3, d Mn2+, e Fe2+, f SO42−

DO and NO3

In the deep flow path, there was significant overlap between the O2 and NO3 reduction zones. When the concentration of oxygen in the infiltrated water was constant and the water flow velocity in the deep flow path was high, the aerobic zone in the aquifer was relatively large, and there was a large DO diffusion distance. In this zone, denitrification occurred under the catalysis of denitrifying bacteria of NO3; aerobic respiration and denitrification occurred synergistically.

Mn2+ and Fe2+

The concentrations of Mn2+ and Fe2+ increased along the flow path, peaking at HJ1-3 and HJ2-3 (17.0 and 200.0 m away from the riverbank, respectively). The concentrations of both ions peaked at shorter distances from the river for the shallow flow path than for the deep flow path. The distance traveled to the monitoring sites with the peak values was longer than that in the shallow flow path for both ions. Moreover, the concentrations remained high near the pumping wells (e.g. HJ3-3), which also differed from the shallow flow path. These differences were mainly the result of deeper monitoring wells and less trapped oxygen being diffused into the aquifer. The seasonal variations in Mn2+ and Fe2+ in the deep groundwater were consistent with those of the shallow groundwater (Fig. 7). The overall pattern was that the hydraulic gradient between the river and groundwater was smaller in winter, and the content of Mn2+ and Fe2+ in groundwater was higher in winter.

SO4 2−

The concentration of SO42− showed a tendency to decrease at first, and then increase with increasing distance from the riverbank. The concentration was below 10.0 mg/L at HJ2-3 (200.0 m away from the riverbank), and the sulfide content reached the peak (about 23.0 × 10−3 mg/L on average) here. Thus, in the deep groundwater, the zone from the riverbank to 200.0 m was the main SO42− reduction zone, and was wider than that in the shallow groundwater.

Variations in the content of some chemical constituents sensitive to redox changes in the vertical direction

The redox zonation indicators varied significantly within the shallow areas (depths less than 10.0 m), whereas variations were less obvious with increasing depth (Fig. 9). From 0 to 3.0 m below the surface, the levels of the electron donor DOC clearly decreased, as did the levels of the electron acceptors DO and NO3. As the reactions progressed, the concentrations of Mn2+ and Fe2+ increased, and reached peak concentrations at 5.0 and 6.0 m below the surface, respectively. At the same depths, the SO42− content decreased.
Fig. 9

Variations in redox zonation indicators with distance of water infiltration in the vertical direction, for the a shallow flow path, and b deep flow path

Redox zonation during bank filtration

Based on analysis of the long sequence of dynamic monitoring data, the threshold concentrations for identifying redox processes in groundwater were formulated according to the hydrodynamic and hydrochemical characteristics of the Liao River water and groundwater in different flow paths (Table 5). The zones were named according to the main redox reaction that primarily reflected the final state in each zone (Fig. 10).
Table 5

Threshold concentrations for identifying redox processes in groundwater

General redox category

Predominant redox process

Water chemistry criteria (mg/L)

O2

NO3

Mn2+

Fe2+

SO42−

Oxic

>1.0

>0.3

<1.0

<7.0

>10.5

Suboxic

O2 reduction

<1.0

>0.3

<1.0

<7.0

>10.5

Anoxic

NO3 reduction

<1.0

<0.3

<1.0

<7.0

>10.5

Mn(IV) reduction

<1.0

<0.3

>1.0

<7.0

>10.5

Fe(III) reduction

<1.0

<0.3

>7.0

>10.5

SO42− reduction

<1.0

<0.3

<10.5

Methanogenesis

<1.0

<0.3

<10.5

Fig. 10

Spatial distribution of redox (reduction) zones in the different flow paths during bank filtration

Vertical differences in the lithology of the aquifer and riverbed-sediment permeability influence water retention time, nutrient and pollutant flux, microbial community structure, acid-base condition, and redox conditions between the river water and groundwater, and have a great influence on biogeochemical processes (Massmann et al. 2004). In the present study, the horizontal extent of the redox zone was different for the different flow paths. The horizontal extent of the redox zone was less than 17.0 m from the riverbank for the shallow flow path, but 200.0 m for the deep flow path. For the shallow flow path, redox zonation showed fuzzier and narrower boundaries, the infiltration rate was slower, the water–rock reaction was more evident, and the superposition phenomenon of the redox zonation was obvious. In addition, for the shallow flow path, the zonation range was narrow and the spatial scales of the reduction zones were about 0–6.5 m for O2/NO3, 6.5–14.0 m for Mn(IV), 14.0–22.0 m for Fe(III), and >22.0 m for SO42− (Fig. 11). The redox zonation in the deep flow path showed more lags and obvious boundaries, and the spatial scales of the reduction zones were 0–12.5 m for O2/NO3, 12.5–100.0 m for Mn(IV), 100.0–240.0 m for Fe(III), and >240.0 m for SO42−. In the nearshore zone, the shallow flow path 10.0 m below the surface was the active redox reaction zone because of precipitation and seasonal river floods. The spatial scales of the near-shore reduction zones were 0–4.0 m for O2/NO3, 4.0–5.4 m for Mn(IV), 5.4–6.8 m for Fe(III), and >6.8 m for SO42− (Fig. 12).
Fig. 11

Redox zonation in the horizontal direction during bank filtration

Fig. 12

Redox zonation in the vertical direction during bank filtration

O2/NO3 reduction zone

In the shallow flow path, aerobic respiration occurred first, and the reaction zone extended 4.0 m horizontally away from the riverbank. Oxygen was almost depleted at this location, and the groundwater environment changed from oxidizing to weakly oxidizing. After infiltration to 6.5 m, denitrification caused the groundwater environment to change from weakly oxidizing to weakly reducing. The O2/NO3 reduction extended a large distance horizontally along the deep flow path, and in the deep flow path the distribution range extended 12.5 m horizontally away from the riverbank. In the vertical direction, the O2/NO3 reduction zone ranged from 0 to 4.0 m below the surface. Both O2 and NO3 oxidize DOC as described in the following reactions:
$$ {\mathrm{CH}}_2\mathrm{O}+{\mathrm{O}}_2\left(\mathrm{aq}\right)\to {\mathrm{CO}}_2+{\mathrm{H}}_2\mathrm{O} $$
(1)
$$ 5{\mathrm{CH}}_2\mathrm{O}+4{{\mathrm{N}\mathrm{O}}_3}^{\hbox{-} }+4{\mathrm{H}}^{+}\to 5{\mathrm{CO}}_2+2{\mathrm{N}}_2+7{\mathrm{H}}_2\mathrm{O} $$
(2)

Mn(IV) reduction zone

In the horizontal direction, the concentration of Mn2+ increased slowly at first, then decreased after peaking at 12.5 m away from the riverbank in the shallow flow path and 17.0 m away from the riverbank in the deep flow path. As the distance to the riverbank increased, the groundwater environment evolved from weakly reducing to reducing, and manganese minerals in the aquifer medium were released into the aquifer by organic complexation or reductive dissolution. In the vertical direction, the peak concentration of Mn2+ was observed at 5.0 m below the surface. At this stage, manganese minerals—expressed as MnO2(s)—were the main electron acceptors in the oxidation of DOC, as expressed in the following reaction:
$$ {\mathrm{CH}}_2\mathrm{O}+2{\mathrm{Mn}\mathrm{O}}_2\left(\mathrm{s}\right)+4{\mathrm{H}}^{+}\to 2{\mathrm{Mn}}^{2+}+3{\mathrm{H}}_2\mathrm{O}+{\mathrm{CO}}_2 $$
(3)

Fe(III) reduction zone

The variation in Fe2+ was similar to that of Mn2+, while the Fe2+ content in the Fe(III) reduction zone peaked at 17.0 m away from the riverbank in the shallow flow path, which was the result of less active hydrodynamic conditions and slower water flow. In the deep flow path, the Fe2+ content peaked at 200.0 m away from the riverbank, whereas in the vertical direction, the peak concentration of Fe2+ was observed at 6.0 m below the surface. During this stage of Fe(III) reduction, iron minerals—expressed as Fe(OH)3(s)—were the main electron acceptors in the oxidation of DOC, and the reaction equation could be expressed as:
$$ {\mathrm{CH}}_2\mathrm{O}+8{\mathrm{H}}^{+}+4\mathrm{Fe}{\left(\mathrm{OH}\right)}_3\left(\mathrm{s}\right)\to 4{\mathrm{Fe}}^{2+}+11{\mathrm{H}}_2\mathrm{O}+{\mathrm{CO}}_2 $$
(4)

SO4 2− reduction zone

In the shallow flow path, SO42− decreased to below the detection limit at about 17.0 m away from the riverbank, and at increasing distances it was no longer detected. The SO42− level was below the detection limit at 200.0 m from the riverbank in the deep flow path, indicating that SO42− continued the redox reaction as an electron acceptor after Fe(III) reduction. In the vertical direction, SO42− was almost below the detection limit at 7.0 m below the surface. The reaction equation can be expressed as follows:
$$ 2{\mathrm{CH}}_2\mathrm{O}+{{\mathrm{SO}}_4}^{2\hbox{-} }+{\mathrm{H}}^{+}\to {\mathrm{H}\mathrm{S}}^{\hbox{-} }+2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{CO}}_2 $$
(5)

Temporal characteristics of redox zonation during bank filtration

The redox zonation was affected by the hydrodynamic conditions, minerals in the aquifer medium, and microorganisms. There was a significant difference in the distribution in different seasons, whereby generally, the distribution of the redox zones was larger in summer than in winter.

In winter, the hydraulic gradient was low, the water infiltration rate was slow, and the water–rock reaction increased. The superposition of redox zonation was clear in the shallow groundwater. The vertical distribution of redox zonation in the deep groundwater was narrower, mainly because the ferromanganese minerals in the medium were fully reduced and released into the groundwater. These minerals migrated toward the center of the regional groundwater depression cone. Because of the small infiltration flux of river water, the vertical dispersion and migration range within the aquifer were small (Fig. 10).

In summer, the river water level was higher, the fluctuation was larger, and the hydrodynamic conditions were stronger. The horizontal and vertical distribution of redox zonation was greater than that in winter. This distribution was the result of the shorter reaction time of water–rock reactions, less iron and manganese being released, and the larger infiltration flux of river water. Consequently, the dispersion and migration range in the aquifer was larger (Fig. 13).
Fig. 13

Spatial distribution of redox (reduction) zones in summer

Release of arsenic from the aquifer medium to groundwater

Arsenic content in the river water was low, but was generally exceedingly high in groundwater (Fig. 14). The groundwater environment has become reductive because of river infiltration. The dissolution of iron or manganese may cause a secondary reaction, and reductive dissolution occurs under the action of organic carbon, resulting in the release and enrichment of heavy metals in mineral surfaces or lattices, and thus the concentration of arsenic increases (Cummings et al. 1999; Guo et al. 2015; Tadanier et al. 2005). The arsenic content increased to 21.5 × 10−3–36.0 × 10−3 mg/L at 17.0 m from the riverbank. This significantly exceeds the drinking-water standards of WHO (≤ 0.01 mg/L). There was a significant positive correlation between arsenic and iron content in the river and groundwater (the linear correlation coefficient was 0.7; Fig. 15). As the river water continued to infiltrate the aquifer, sulfate reduction occurred and the resulting HS and Fe2+ formed sulfides (such as FeS). This resulted in the adsorption or co-precipitation of arsenic, thus reducing the arsenic in the groundwater (Kocar et al. 2010; Saalfield and Bostick 2009). The arsenic content was reduced to about 0.01 mg/L at 350.0 m from the riverbank, near the pumping wells.
Fig. 14

Boxplots of arsenic content in river water and groundwater

Fig. 15

Correlation of iron and arsenic content in river water and groundwater

Conclusions

During bank filtration, organic matter was the main electron donor, and electrons reacted with acceptors in the following order based on energy: O2, NO3, Mn(IV) and Fe(III) oxides or hydroxides, SO42−, and then CO2. The horizontal extent of the redox zone was controlled by the permeability of riverbed sediments and the aquifer at different depths. The redox zone extended horizontally to 17.0 m from the riverbank in the case of the shallow flow path, and 200.0 m in the case of the deep flow path. The vertical extent of the redox zone was affected by precipitation and seasonal river floods and ranged from 0 to 10.0 m below the surface. Arsenic was released to the groundwater with the reduction of iron and manganese oxides or hydroxides; however, bank filtration and the SO42− reduction zone can remove arsenic, leading to potable water being available to meet water demand. The results obtained in this study will help in management plans that address drinking-water quality.

Notes

Acknowledgements

The authors would like to thank the editors of Hydrogeology Journal and the reviewers for their thoughtful and constructive comments, which helped improve the manuscript.

Funding Information

This work was supported by the National Natural Science Foundation of China (Grant numbers: 41372238, 41402209, 41502223, 41602271, and DD20160207).

References

  1. Appelo CAJ, Postma D (2005) Geochemistry, groundwater and pollution, 2nd edn. Balkema, Leiden, The NetherlandsCrossRefGoogle Scholar
  2. Bertelkamp C, Verliefde ARD, Schoutteten K, Vanhaecke L, Vanden Bussche J, Singhal N, van der Hoek JP (2016) The effect of redox conditions and adaptation time on organic micropollutant removal during river bank filtration: a laboratory-scale column study. Sci Total Environ 544:309–318.  https://doi.org/10.1016/j.scitotenv.2015.11.035 CrossRefGoogle Scholar
  3. Bourg ACM, Bertin C (1993) Biogeochemical processes during the infiltration of river water into an alluvial aquifer. Environ Sci Technol 27:661–666.  https://doi.org/10.1021/es00041a009 CrossRefGoogle Scholar
  4. Bouwer H (2002) Artificial recharge of groundwater: hydrogeology and engineering. Hydrogeol J 10:121–142.  https://doi.org/10.1007/s10040-001-0182-4 CrossRefGoogle Scholar
  5. Brown CJ, Schoonen MAA, Candela JL (2000) Geochemical modeling of iron, sulfur, oxygen and carbon in a coastal plain aquifer. J Hydrol 237:147–168.  https://doi.org/10.1016/S0022-1694(00)00296-1 CrossRefGoogle Scholar
  6. China GB (2006) National Standard of the People’s Republic of China: standards for drinking water quality (GB5749–2006) (in Chinese). National Health and Family Planning Commission of the People’s Republic of China, BeijingGoogle Scholar
  7. Cummings DE, Frank Caccavo JR, Fendorf S, Rosenzweig RF (1999) Arsenic mobilization by the dissimilatory Fe(III)-reducing bacterium Shewanella alga BrY. Environ Sci Technol 33:723–729.  https://doi.org/10.1021/es980541c CrossRefGoogle Scholar
  8. Diem S, Cirpka OA, Schirmer M (2013) Modeling the dynamics of oxygen consumption upon riverbank filtration by a stochastic-convective approach. J Hydrol 505:352–363.  https://doi.org/10.1016/j.jhydrol.2013.10.015 CrossRefGoogle Scholar
  9. European Commission (1998) Guidelines of water intended for human consumption. Off J Eur Communities, EC, BrusselsGoogle Scholar
  10. Farnsworth CE, Hering JG (2011) Inorganic geochemistry and redox dynamics in bank filtration settings. Environ Sci Technol 45:5079–5087.  https://doi.org/10.1021/es2001612 CrossRefGoogle Scholar
  11. Gandy CJ, Smith JWN, Jarvis AP (2007) Attenuation of mining-derived pollutants in the hyporheic zone: a review. Sci Total Environ 373:435–446.  https://doi.org/10.1016/j.scitotenv.2006.11.004 CrossRefGoogle Scholar
  12. Greskowiak J, Prommer H, Massmann G, Nützmann G (2006) Modeling seasonal redox dynamics and the corresponding fate of the pharmaceutical residue phenazone during artificial recharge of groundwater. Environ Sci Technol 40:6615–6621.  https://doi.org/10.1021/es052506t CrossRefGoogle Scholar
  13. Guo H, Liu Z, Ding S, Hao C, Xiu W, Hou W (2015) Arsenate reduction and mobilization in the presence of indigenous aerobic bacteria obtained from high arsenic aquifers of the Hetao basin, Inner Mongolia. Environ Pollut 203:50–59.  https://doi.org/10.1016/j.envpol.2015.03.034 CrossRefGoogle Scholar
  14. Hamann E, Stuyfzand PJ, Greskowiak J, Timmer H, Massmann G (2016) The fate of organic micropollutants during long-term/long-distance river bank filtration. Sci Total Environ 545–546:629–640.  https://doi.org/10.1016/j.scitotenv.2015.12.057 CrossRefGoogle Scholar
  15. Hancock PJ (2002) Human impacts on the stream-groundwater exchange zone. Environ Manag 29(6):763–781.  https://doi.org/10.1007/s00267-001-0064-5 CrossRefGoogle Scholar
  16. Haque S, Johannesson KH (2006) Arsenic concentrations and speciation along a groundwater flow path: the Carrizo Sand Aquifer, Texas, USA. Chem Geol 228:57–71.  https://doi.org/10.1016/j.chemgeo.2005.11.019 CrossRefGoogle Scholar
  17. Heberer T, Massmann G, Fanck B, Taute T, Dünnbier U (2008) Behaviour and redox sensitivity of antimicrobial residues during bank filtration. Chemosphere 73:451–460.  https://doi.org/10.1016/j.chemosphere.2008.06.056 CrossRefGoogle Scholar
  18. Henzler AF, Greskowiak J, Massmann G (2016) Seasonality of temperatures and redox zonations during bank filtration: a modeling approach. J Hydrol 535:282–292.  https://doi.org/10.1016/j.jhydrol.2016.01.044 CrossRefGoogle Scholar
  19. Hiscock KM, Grischek T (2002) Attenuation of groundwater pollution by bank filtration. J Hydrol 266:139–144.  https://doi.org/10.1016/S0022-1694(02)00158-0 CrossRefGoogle Scholar
  20. Hu B, Teng Y, Zhai Y, Zuo R, Li J, Chen H (2016) Riverbank filtration in China: a review and perspective. J Hydrol 541:914–927.  https://doi.org/10.1016/j.jhydrol.2016.08.004 CrossRefGoogle Scholar
  21. Huntscha S, Rodriguez Velosa DM, Schroth MH, Hollender J (2013) Degradation of polar organic micropollutants during riverbank filtration: complementary results from spatiotemporal sampling and push-pull tests. Environ Sci Technol 47:11512–11521.  https://doi.org/10.1021/es401802z CrossRefGoogle Scholar
  22. Jung HB, Zheng Y, Rahman MW, Rahman MM, Ahmed KM (2015) Redox zonation and oscillation in the hyporheic zone of the Ganges-Brahmaputra-Meghna Delta: implications for the fate of groundwater arsenic during discharge. Appl Geochem 63:647–660.  https://doi.org/10.1016/j.apgeochem.2015.09.001 CrossRefGoogle Scholar
  23. Kedziorek MAM, Bourg ACM (2009) Electron trapping capacity of dissolved oxygen and nitrate to evaluate Mn and Fe reductive dissolution in alluvial aquifers during riverbank filtration. J Hydrol 365:74–78.  https://doi.org/10.1016/j.jhydrol.2008.11.020 CrossRefGoogle Scholar
  24. Kedziorek MAM, Geoffriau S, Bourg ACM (2008) Organic matter and modeling redox reactions during river bank filtration in an alluvial aquifer of the Lot River, France. Environ Sci Technol 42:2793–2798.  https://doi.org/10.1021/es702411t CrossRefGoogle Scholar
  25. Kocar BD, Borch T, Fendorf S (2010) Arsenic repartitioning during biogenic sulfidization and transformation of ferrihydrite. Geochim Cosmochim Acta 74:980–994.  https://doi.org/10.1016/j.gca.2009.10.023 CrossRefGoogle Scholar
  26. Lewandowski J, Putschew A, Schwesig D, Neumann C, Radke M (2011) Fate of organic micropollutants in the hyporheic zone of a eutrophic lowland stream: results of a preliminary field study. Sci Total Environ 409:1824–1835.  https://doi.org/10.1016/j.scitotenv.2011.01.028 CrossRefGoogle Scholar
  27. Lovley DR, Phillips EJP (1988) Novel mode of microbial energy metabolism: organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl Environ Microbiol 54(6):1472–1480.  https://doi.org/10.1103/PhysRevLett.50.1998 Google Scholar
  28. Massmann G, Pekdeger A, Merz C (2004) Redox processes in the Oderbruch polder groundwater flow system in Germany. Appl Geochem 19:863–886.  https://doi.org/10.1016/j.apgeochem.2003.11.006 CrossRefGoogle Scholar
  29. Massmann G, Greskowiak J, Dünnbier U, Zuehlke S, Knappe A, Pekdeger A (2006) The impact of variable temperatures on the redox conditions and the behaviour of pharmaceutical residues during artificial recharge. J Hydrol 328:141–156.  https://doi.org/10.1016/j.jhydrol.2005.12.009 CrossRefGoogle Scholar
  30. Massmann G, Nogeitzig A, Taute T, Pekdeger A (2008a) Seasonal and spatial distribution of redox zones during lake bank filtration in Berlin, Germany. Environ Geol 54:53–65.  https://doi.org/10.1007/s00254-007-0792-9 CrossRefGoogle Scholar
  31. Massmann G, Dünnbier U, Heberer T, Taute T (2008b) Behaviour and redox sensitivity of pharmaceutical residues during bank filtration: investigation of residues of phenazone-type analgesics. Chemosphere 71:1476–1485.  https://doi.org/10.1016/j.chemosphere.2007.12.017 CrossRefGoogle Scholar
  32. Polomčić D, Hajdin B, Stevanović Z, Bajić D, Hajdin K (2013) Groundwater management by riverbank filtration and an infiltration channel: the case of Obrenovac, Serbia. Hydrogeol J 21:1519–1530.  https://doi.org/10.1007/s10040-013-1025-9 CrossRefGoogle Scholar
  33. Prommer H, Stuyfzand PJ (2005) Identification of temperature-dependent water quality changes during a deep well injection experiment in a pyritic aquifer. Environ Sci Technol 39:2200–2209.  https://doi.org/10.1021/es0486768 CrossRefGoogle Scholar
  34. Saalfield SL, Bostick BC (2009) Changes in iron, sulfur, and arsenic speciation associated with bacterial sulfate reduction in ferrihydrite-rich systems. Environ Sci Technol 43:8787–8793.  https://doi.org/10.1021/es901651k CrossRefGoogle Scholar
  35. Scow KM, Hicks KA (2005) Natural attenuation and enhanced bioremediation of organic contaminants in groundwater. Curr Opin Biotechnol 16:246–253.  https://doi.org/10.1016/j.copbio.2005.03.009 CrossRefGoogle Scholar
  36. Sharma L, Greskowiak J, Ray C, Eckert P, Prommer H (2012) Elucidating temperature effects on seasonal variations of biogeochemical turnover rates during riverbank filtration. J Hydrol 428-429:104–115.  https://doi.org/10.1016/j.jhydrol.2012.01.028 CrossRefGoogle Scholar
  37. Stuyfzand PJ (2011) Hydrogeochemical processes during riverbank filtration and artificial recharge of polluted surface waters: zonation, identification, and quantification. In: Shamrukh M (ed) Riverbank filtration for water security in desert countries. Springer, Dordrecht, The Netherlands, pp 97–128CrossRefGoogle Scholar
  38. Su X, Lu S, Gao R, Su D, Yuan W, Dai Z, Papavasilopoulos EN (2017) Groundwater flow path determination during riverbank filtration affected by groundwater exploitation: a case study of Liao River, Northeast China. Hydrol Sci J 62(14):2331–2347.  https://doi.org/10.1080/02626667.2017.1383609 CrossRefGoogle Scholar
  39. Tadanier CJ, Schreiber ME, Roller JW (2005) Arsenic mobilization through microbially mediated deflocculation of ferrihydrite. Environ Sci Technol 39:3061–3068.  https://doi.org/10.1021/es048206d CrossRefGoogle Scholar
  40. Tufenkji N, Ryan JN, Elimelech M (2002) Peer reviewed: the promise of bank filtration. Environ Sci Technol 36(21):422A–428A.  https://doi.org/10.1021/es022441j CrossRefGoogle Scholar
  41. USEPA (2011) 2011 edition of the drinking water standards and health advisories. Office of Water, US Environmental Protection Agency, Washington, DCGoogle Scholar
  42. von Rohr MR, Hering JG, Kohler H-PE, von Gunten U (2014) Column studies to assess the effects of climate variables on redox processes during riverbank filtration. Water Res 61:263–275.  https://doi.org/10.1016/j.watres.2014.05.018 CrossRefGoogle Scholar
  43. WHO (2011) Guidelines for drinking-water quality, 4th edn. World Health Organization, GenevaGoogle Scholar
  44. Worch E, Grischek T, Börnick H, Eppinger P (2002) Laboratory tests for simulating attenuation processes of aromatic amines in riverbank filtration. J Hydrol 266:259–268.  https://doi.org/10.1016/S0022-1694(02)00169-5 CrossRefGoogle Scholar
  45. Xie X, Wang Y, Ellis A, Su C, Li J, Li M, Duan M (2013) Delineation of groundwater flow paths using hydrochemical and strontium isotope composition: a case study in high arsenic aquifer systems of the Datong basin, northern China. J Hydrol 476:87–96.  https://doi.org/10.1016/j.jhydrol.2012.10.016 CrossRefGoogle Scholar
  46. Yuan W (2017) Biogeochemical process of Fe and Mn during river bank infiltration affected by groundwater exploiting (in Chinese). PhD Thesis, Jilin University, Changchun, ChinaGoogle Scholar
  47. Zhang L (2015) Study on pore water hydrogeochemistry evolution of riverbed sedimentation zone under condition of riverbank filtration: an example in Shenyang Huangjia water source (in Chinese). MSc Thesis, Jilin University, Changchun, ChinaGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Xiaosi Su
    • 1
    • 2
  • Shuai Lu
    • 1
    • 3
  • Wenzhen Yuan
    • 4
  • Nam Chil Woo
    • 5
  • Zhenxue Dai
    • 1
    • 2
  • Weihong Dong
    • 1
    • 3
  • Shanghai Du
    • 1
    • 2
  • Xinyue Zhang
    • 3
  1. 1.Institute of Water Resources and EnvironmentJilin UniversityChangchunChina
  2. 2.College of Construction EngineeringJilin UniversityChangchunChina
  3. 3.College of Environment and ResourcesJilin UniversityChangchunChina
  4. 4.Chinese Academy of Geological SciencesBeijingChina
  5. 5.Department of Earth System SciencesYonsei UniversitySeoulSouth Korea

Personalised recommendations