Environ Earth Sci DOI 10.1007/s12665-014-3637-3

ORIGINAL ARTICLE

Role of shallow alluvial stratigraphy and Holocene geomorphology on groundwater arsenic contamination in the Middle Ganga Plain, India Sudarsan Sahu • Dipankar Saha

Received: 13 November 2013 / Accepted: 19 August 2014 Ó Springer-Verlag Berlin Heidelberg 2014

Abstract Floodplain geomorphology and Late Quaternary alluvial stratigraphy influence groundwater arsenic contamination in the Gangetic Plains in India. Elevated concentrations of arsenic ([50 lg/L) in the Middle Ganga Plain (MGP) are spatially related with dark grey to black coloured organic-rich clay sediments found both in the Active (AFP, T0-surface) and Older Floodplains (OFP, T1surface) of the Ganga River. The floodplains comprise Newer Alluvium of Holocene age, mainly derived from Himalaya. Abandoned/palaeochannel cut-offs, either filled (often in OFP) or in the process of filling under fluviolacustrine environments (often in AFP) host such clay deposits. The settlements on anthropogenic fills (to avoid flood inundation) over such palaeochannels exhibit higher concentrations of arsenic in groundwater (max. 987 lg/L). The organic carbon from the clay bodies when released to groundwater promotes reducing environment that causes release of arsenic entrapped in the sediments. The sandy areas, such as scroll bar ridges, levees and point bar platforms, exhibit low concentrations of arsenic. The model on groundwater arsenic distribution and Holocene geomorphic units can be used for micro-level delineation of arsenic-free shallow aquifers for rural water supply in the arseniccontaminated areas. Keywords Groundwater  Arsenic contamination  Gangetic Plains  India  Quaternary geology  Geomorphology  Organic carbon

S. Sahu (&)  D. Saha Central Ground Water Board, MER, Patna, India e-mail: [email protected]

Introduction Groundwater arsenic contamination is a major health hazard in the lowlands of fluvial and fluvio-deltaic tracts in different parts of the world (Berg et al. 2001; Gigera et al. 2003; Polya et al. 2005; Acharya and Shah 2007; Burgess et al. 2010; McArthur et al. 2012). The most widespread contamination is reported from Bengal Delta Plain (BDP), covering parts of West Bengal, India and Bangladesh, affecting more than 50 million inhabitants (Nickson et al. 1998; BGS and DPHE 2001; Sharma et al. 2006; Johnston et al. 2010; Burgess et al. 2010). In recent years, high incidence of arsenic in groundwater has been reported from the Middle Ganga Plain (MGP), covering parts of Bihar and Uttar Pradesh in India too (Chakraborty et al. 2003; Acharya 2004; Shah 2008; Nayak et al. 2008; Saha et al. 2010; Kumar et al. 2010). Significant achievements have been made on understanding the role of geomorphology in groundwater arsenic contamination in fluvio-deltaic set-up of BDP (e.g. McArthur et al. 2004; Weinman et al. 2008; Hoque et al. 2009; Neumann et al. 2010; Mukherjee et al. 2012; Sahu and Shukla 2013; Malik and Biswas 2014). Spatial heterogeneity in distribution of arsenic concentration in groundwater has been documented (Sengupta et al. 2004). Van Geen et al. (2003) have documented 50-fold increase in arsenic concentrations in wells located several metres apart and tapping aquifers of same depth at villages located in the lower reaches of BDP in Bangladesh. Such localised variability in groundwater arsenic has been linked with geomorphic units and near-surface geology (Khan and Hoque 2002; Hoque et al. 2009). Weinman et al. (2008) have documented lower arsenic concentrations in hand pumps located on levees and bars comprising permeable sand, representing elevated parts of the floodplain in

123

Environ Earth Sci

Bangladesh. In contrast, the authors report higher concentrations of groundwater arsenic in hand pumps located on meandering stream levees with low elevation and thick sequences of finer sediments. Since the reporting of elevated concentration of arsenic in MGP, the researches carried out so far have focussed on the distribution of arsenic in different aquifers and chemistry of arsenic-rich groundwater (CGWB and PHED 2005; Acharya and Shah 2007; Saha 2009, 2010, 2011a; Raju 2012). Attempts have also been made to understand the recharge mechanism of arsenic-contaminated aquifers through application of stable oxygen–deuterium isotopes (Saha et al. 2011b). However, no attempt has so far been made on linking the arsenic distribution, geomorphology and near-surface stratigraphy in the MGP. Unlike BDP, where sea level change has influenced the sedimentation history during Late Quaternary; the MGP, located at a distance of *700 km from the present coastline, is marked with fluvio-lacustrine depositional environment. The contaminated areas typically fall in the low-lying floodplains of the Ganga River comprising the Newer Alluvium of Holocene age. In Bihar State located in MGP, the contamination is reported from a corridor of \1–20 km wide along the course of the Ganga affecting both the banks. Considering regulating limit of 50 lg/L of arsenic in groundwater, the contaminated areas are spread over *9,000 km2 (9.5 % of the geographical area of the state) with *10 million people (9.6 % of state populace) residing in the risk zone (Sahu and Dwivedi 2012). If the 10 lg/L limit, as recommended by WHO (1993), is considered, the affected area and population will increase significantly. The present investigation deals with the study of geomorphic features and shallow alluvial stratigraphy vis-a-vis the pattern of groundwater arsenic distribution in MGP. The role of Quaternary stratigraphy on the groundwater arsenic contamination in the region has been studied earlier (Shah 2008), where a broad relation between arsenic distribution and morphostratigraphy has been attempted. The present study aims at understanding the variability of arsenic in groundwater in different geomorphic sub-environments with characteristic sedimentology and stratigraphic set-ups. Other than identifying the vulnerable environments for groundwater arsenic, those sub-environments have also been delineated which host the local-scale aquifers with lower concentrations of arsenic in groundwater.

Study area set-up The area, located in the Bihar State, forms a part of the MGP, situated in the middle part of the east–west elongated Gangetic Plains. The area covers *535 km2, falling

123

in parts of Bhojpur and Buxar districts of the state, located in the South Ganga Plain (SGP), a geomorphic unit of MGP (Govt. of Bihar 2007) (Fig. 1). The area extends between the eastern longitudes of 84°340 0000 and 84°130 2700 and the northern latitudes of 25°340 5200 N and 25°450 2000 . Geographically, the area forms a part of the 5–20 km wide floodplain of Ganga, attached to its southern bank. The floodplain has been formed by north/northwestward migration of the river due to tilting in this part of the basin during Late Quaternary (Saha et al. 2010; Sahu and Saha 2013). Within the area, the Ganga River has a lower gradient (0.005–0.007 %) and the sediment load consists mainly of fine to very fine sand with admixtures of silt and mud. In this low-energy hydrological set-up, though braid bars are formed within the meandering course of the river, those are visible only during low-flow stages. Towards south, the study area merges with the upland Pleistocene surface, which remains beyond the reach of Ganga flood water. The highlands of Proterozoic Vindhyan rocks of Indian peninsula (the craton) are situated further south (Fig. 1), which extend northward below the Quaternary deposits, forming the basement in this part of MGP (Om Prakash et al. 1990; Saha et al. 2007). Morphostratigraphically, the upland alluvial surface is referred as Older Alluvium of Middle to Upper Pleistocene period (Chakraborty and Chattopadhyay 2001). The age has been assigned on the basis of landform slope, degree of dissection, degree of pedogenic transformations, state of preservation of landform elements and the nature and degree of oxidation of the constituent sediments in the morphostratigraphic unit (Dayal 1997). The river valleys are incised on to the upland surface with the deposition of Newer Alluvium during the early Holocene to Recent (Om Prakash et al. 1990). Within the study area, the Newer Alluvium represents two terrace deposits; (1) the upper terrace forming the Older Floodplain (OFP) remains beyond the normal flood events, (2) the lower terrace, which represents the Active Floodplain (AFP), has been deposited on the entrenched surface upon OFP and is prone to regular flooding during monsoon. The arsenic-contaminated aquifers are mainly confined within the shallow depth of *30 m below ground within the Newer Alluvium (Saha et al. 2011a; Sahu 2013). The Himalaya has been indicated as the source of arsenic-rich sediments (Acharya 2004; Shah 2008). The sediments comprise fine to very fine micaceous grey to dark grey sand and mud, which are in an un-oxidised state (Saha et al. 2011a, 2014; Sahu 2013). In contrast, the Older Alluvial sediments derived from peninsular highlands are oxidised, yellow to brownish-yellow in colour, coarse grained and mixed with kankars (carbonate nodules) (Saha et al. 2008, 2009, 2011a, 2014; Sahu 2013). In a similar trend as reported from BDP (Ravenscroft et al. 2005; Von

Environ Earth Sci

A

C

B

D

Fig. 1 Location map of the study area. Inset a shows the Ganga Basin in India. Inset b The Bihar state with the north and south Ganga plains. The Sone-Ganga interfluve is situated at the south-western parts of the state. The southern part of the state is covered by Indian peninsula. Inset c depicts the boundary between the newer (I) and the older alluvium (II). The hatched part of c represents the study area.

The geomorphology and Quaternary morphostratigraphic map of the study area with the details of the observation points made have been produced in inset d (traced from satellite map). The area of the rectangular inset in 1d has been studied intensively for abundance of arsenic in groundwater (produced as Fig. 5)

Bro¨mssen et al. 2007), the Older Alluvium in MGP also shows lower arsenic concentrations in groundwater (Shah 2008, 2009, 2010). The arsenic contamination in the study area in MGP is marked with wide spatial variation. The survey by Chakraborty et al. (2003) in two villages (Bariswan and Semaria-Ojhapatti) reported arsenic level in groundwater beyond 50 lg/L in *59 % of hand pumps tapping the shallow aquifer.

of India topographic sheets and satellite data, in conjunction with the field survey results. The methodology adopted is tabulated below; 1.

2.

3. Methodology The study involved data collection and generation from field survey, and analysis of sediment and water samples. Besides, various thematic maps were prepared from Survey

4.

Remote sensing investigation: Interpretation of the main geomorphologic and the morphostratigraphic units by satellite image interpretation. Litho-stratigraphic investigation: study of sub-surface alluvial exposures and borehole lithologs, sampling of sediments for analysis of grain size and organic carbon content. Geochemical study: quantitative analyses of arsenic content in groundwater samples collected from hand pumps. Integrated morpho-hydrogeologic analysis: overlaying the groundwater arsenic concentrations on geomorphic features and investigating its relation with stratigraphy.

123

Environ Earth Sci

Satellite data with ground resolution of 21.0 m (IRS ID, sensor LISS III) were also interpreted for May (pre-monsoon) and November (post-monsoon) coupled with selective ground truth verifications. Topographical maps were compared with the satellite maps (year 2009) to assess the channel movement rates of Ganga and the temporal character of the water bodies in the floodplain. For identification of various sedimentary facies and stratigraphic sequences, seven trenches (2–3 m deep and 1.5–2.7 km long) excavated for laying telephone cables and 14 sand mines (4–7 m deep) existing in the area were studied. Drill core samples up to 30 m depth from 15 bore holes (Fig. 1) were obtained to prepare lithological sections up to centimeter scale. Sediment samples (n = 10) were collected from the sand mine and trench heads for analysis of particle size distribution in different sedimentary facies. The analysis was carried out using standard set of sieves of sizes 2,000, 1,000, 500, 250, 180, 125 and 70 lm. Sediment fractions of size \70 lm were analysed by hygrometer. The very fine sand, silt and mud samples were also subjected to the analysis of organic carbon content in them. For the purpose, additional 15 black mud samples (locally known as kali mitti) were collected from 5 pits (1 9 1 9 1 m size), specially dug in the low-lying areas during the investigation (Fig. 1). Groundwater samples were collected from 285 hand pumps (depth 10–30 m) distributed in different geomorphologic units. Each sample was collected after 10 min of purging to get fresh aquifer yield. The samples were filtered on spot (using 41 no. Whatman filter paper) and were acidified to bring the pH level at 2. The arsenic concentration was analysed in the laboratory of Indian School of Mines, Dhanbad using AAS with graphite furnace hydride generator (Shimadzu 6300) with detection limit at 2 lg/L.

Morphology of floodplain and sedimentation patterns The floodplain in the area exhibits a monotonously flat landscape with altitude varying between 55 and 62 m above mean sea level (msl). Abandoned channels and a series of meander scars are left behind in the floodplain in course of migration of the Ganga. The geomorphologic sub-features, briefed below, include the low-lying mudflats, channel cut-off lakes, levees, and point bars with typical ridge (scroll bar) and swale topography. Channel cut-offs and mudflats The cut-offs as well as small abandoned cross-bar channels associated with the meander scars, mainly in the OFP, have been levelled with finer argillaceous clayey sediments. The sediments are deposited by flood water and also

123

contributed by the vertical accretion of silt/clay and organic matter from stagnant water under lacustrine environment. This is exemplified by two existing lakes in the AFP (Fig. 1). Both ends of these cut-off lakes are blocked by sediments, resulting in ponded water sedimentation. Large angle of divergence between the stem channel and the cutoff channel hinders the bed load entry from the main channel into the lakes. In many cases, the abandoned crossbar channels also form water bodies (Fig. 1d), holding water for number of years. A comparison of toposheet of the year 1968 with the satellite data of 2009 revealed that such water bodies also form potential sites of biomass accumulation. The study of satellite data for pre- and postmonsoon seasons indicated that the filled up old channels in OFP still form natural depressions (lying 2–3 m below the adjacent scroll bar ridges) and remain under seasonal floodwater (Fig. 1). When flood water recedes, those behave as mudflats. Point bars: the ridge and swale topography The point bar surfaces show alternating ridges and swales or depressions (meander scroll topography), formed as a result of channel shifting and formation of small and large cross-bar channels. The distribution pattern of swales on point bars indicates continuous migration of meanders. A comparison between the courses of the Ganga River of the year 1922 (Singh and Singh 1971) and that of 2009 revealed that the meander channel migrated at an average rate of 50–60 m/year. Thus, each swale might have witnessed several floods. The active point bars exhibit cross-bar channels with widths ranging within 25–200 m. Meander lobe cut-offs originating from channel diversion along a cross-bar channel indicate that the meander cut-offs are generally ‘‘chute cut-offs’’ as defined by Allen (1965). Except few cross-bar channels, which bear minor discharges throughout the year, the distant ones get water only during the high-flow stages of the Ganga. As the meander migrates, those are abandoned and become draped by fine-grained muddy sediments. Levees Levees in the study area are wide and subdued in comparison to those observed in the downstream parts particularly around Patna urban area (Saha et al. 2010). They dip gently towards south and merge with the floodplain. The channel meandering and migration of the Ganga towards north/northwestward might have resulted in such less prominent levees. Paucity of coarse suspended sediment in the flood water is also one of the reasons for such subdued levees (Hudson and Heitmuller 2003). The levees along the

Environ Earth Sci

cut-bank of the old meander scars are composed of silt and mud with infrequent lenses of very fine sand. However, the levees on the inner side of the meanders are relatively rich in sand with minor lenses of mud.

Shallow alluvial stratigraphy Ten sedimentary facies (Table 1) have been identified within a depth of 30 m below ground level (bgl). Those can be grouped into six broad categories; (a) facies l1 and l2, representing channel lag deposits with carbonate nodules and sand mostly in 1,000–3,000 lm size range, (b) facies f1 and f2, representing a well-sorted fine sand with proportions of mica flakes, where *75–85 % of the particles are in 130–240 lm size range, (c) facies f3, a moderately well-sorted very fine sand with varying proportions of silt and 70–85 % of the grains falling in the range of 92–110 lm, (d) facies f4, which is a clean, fine to coarse silt with a proportion of very fine sand and 69–82 % of the particles in the size range of 22–55 lm (e) facies m1 and m2, representing mud with varying proportions of silt and clay, and (f) facies c1 and c2, representing the black clay and organic matter with minor silt. There is a widespread drape of silt/mud and clay (facies f4, m1, m2, c1 and c2) over the grey coloured shallow aquifer sand (SAS). The SAS, overlying the lag deposits (facies l1 and l2), locally hosts the arsenic-rich groundwater. The SAS is 6–15 m thick, comprising the sedimentary facies f1 and f2, which represent bar formations and transitory sand in channels (Table 1). Often these sand bodies are interspersed with thin layers of mud, deposited during the high stages of floods. Seven stratigraphic sequences (strat. seq. S1–S7) have been identified within 30 m bgl (Table 2, Fig. 2). The thickness and the nature of sedimentary facies overlying the SAS differentiate these sequences. The sequences have evolved in distinct fluvio-lacustrine environments such as channel cut-off lakes, scroll bars, point bar platforms and intervening floodplains. The stratigraphic sequences (Fig. 2) exhibit fining upward character with SAS at the bottom. The SAS grades upward to (1) thin floodplain mud (strat. seq. S1), (2) silt followed by mud (strat. seq. S2), (3) very fine sand to coarse silt and black clay followed by mud (strat. seq. S3), (4) two fining upward sequences terminated by black clay/ mud (strat. seq. S4), (5) thick (7–15 m) black clay/mud, rich in organic matter, followed by floodplain mud (strat. seq. S5) (6) similar to S5 and 2–3 m of reworked floodplain mud/clay as the anthropogenic fill at the top (strat. seq. S6), and (7) interlayer of silt/mud/clay with minor sand lenses (strat. seq. S7).

The stratigraphic sequence S1 forms the topographic ridges with 2–3 m elevation above the adjacent floodplain. Thickness of the silt/mud cover at the top of the sequence is low, varying within 0.50–0.80 m. It increases to 2–4 m in S2 representing point bar platform sequence, where very fine sand and silt components predominate within the mud cover. These two sequences S1 and S2 are widely distributed in areas of habitation. The sequence S3 is similar to S2, except an organic-rich mud (0.5–1.5 m thick) observed in between the SAS at the bottom and the silt/mud cover at the top. This sequence has evolved in small and large cross-bar channels. The sequence S4 represents repeated cross-bar channel deposits, indicating renewed floodwater activity. The sequence S5 comes next to S1 in aerial extent, exhibiting curvilinear tracts of palaeochannels, ranging 0.85–1.80 km in width, similar to the present day width of the Ganga River. The areas with S5 represent sequences with thick (channel deep 7–15 m) finer argillaceous sediment, overlying the SAS. Being the sites of palaeo-lakes, these areas possess relatively thick organic-rich black mud (facies c1 and c2) other than the floodplain silt and mud (facies m1 and m2). The areas under S5 are flat and low lying, and pave the way for floodwater into the plain. The sequence S6 is similar to S5 except an anthropogenic fill of 2–3 m at the top to keep the settlements beyond the regular flood levels. The sequence S7 consists of very thin alternation of silt/mud/clay and clean sand layers. It is restricted to the adjoining areas of palaeochannels depicting proximal floodplain sequence.

Distribution of arsenic Groundwater samples were collected from villages settled on various parts of the floodplain representing the stratigraphic sequences S1, S2, S3, S4, S6 and S7. The sequence S5 represents low-lying landform and could not be sampled due to non-availability of any settlement over it. This sequence is occupied mostly by paddy fields and none of the irrigation tube wells were found to be running during the groundwater sampling period. In the study area, arsenic concentrations vary widely between below detection limit (BDL) and *1,000 lg/L, with 34 % of the samples showing values exceeding the limit of 50 lg/L. The arsenic distribution pattern, however, exhibits a relation with the major geomorphologic features in the floodplain (Fig. 3). The intensity of contamination in different stratigraphic sequences can be classified into three categories; (1) Low, with arsenic in groundwater largely\50 lg/L (range BDL-141 lg/L) in the sequences S1, S2 and S7 (2) Moderate, with arsenic frequently exceeding 50 lg/L (range BDL-342 lg/L) in the sequences S3 and S4, and (3) High, with extensive incidence of higher concentrations of arsenic [50 lg/L (range 9–987 lg/L) in the sequence S6 (Fig. 4).

123

123

Silt/clay. Basically flood plain mud deposits, with higher content of transgenic clay

m1

m2

(e)

(f)

Silt and clay. Basically flood plain mud deposits, with higher content of silt

f4

Clay rich with organic matter, with minor silt

Organic black clay with minor or no silt

c1

c2

Silt/mud with a good proportion of very fine sand

f3

(d)

Fine sand with a dominating proportion (*70–80 %) of mica flakes Very fine sand with a proportion of silt/mud

f2

(c)

Fine sand with mica flakes. Basically transitory sand bodies on channels

f1

Medium to coarse sand with carbonate nodules and infrequent iron flakes and nodules

l2

(b)

Coarse to gravel size reworked carbonate nodules

l1

(a)

Lithology

Facies

Sl. no.

Table 1 Sedimentary facies identified in the area

Greyish black to black

Light black (locally termed as kalli mitti)

Light yellow to grey

Brown to light yellow

Brown

Grey, shining dark grey (when biotite predominates) Light grey to Brown

Grey

Light yellow to yellowish brown

Grey

Colour

1:10:89

3:25:72

5:25:70

5:60:35

14:69:14

92:6:2

93–98:7–2:0

100:0:0

–

Sand: silt: clay (% ratio)

4

12

15

31

52

91

107

210

595

–

Average mean grain size (lm)

5.9–6.6

3.2–6.1

0.8–1.6

0.6–1.2

0.1–0.6

0.1–0.8

–

–

–

Organic Carbon content in (g/kg)

Lacustrine deposits

Over bank flood plain deposits

Low-energy channel deposits (cross-bar channels)/over bank deposits (upper reaches of point bars)

Channel sand as bar deposits

Channel sand as bar deposits (braid bar and lower reaches of point bars) and other transitory sand on channel bed

Channel lag deposits of Ganga/ channel deposit of Sone River

Interpreted depositional setting

Environ Earth Sci

Lithology

Clean grey sands, horizontally bedded, sometimes cross bedded. Fining upward and capped by thin mud layers

Considerably thick silt/very fine sand followed upward by mud. Underlain by clean grey sand

Similar to S2, except that a black organic-rich layer below the top mud cover. Underlain by clean grey sand

Similar to S2, except that a second cycle of fining upward sequence continues, finally capped by black mud (organic rich). Underlain by clean grey sand

Thick, organic-rich mud (dark grey to black colour) with thin inter-layers of sand and silt, and overlying flood plain mud of limited thickness. Occasionally it directly overlies deeper channel-lag deposit with or without minor SAS

S5 with artificial fill at the top

Mud with interlayers of fine sand, silt and clay lenses

Stratigraphic sequences

S1

S2

S3

S4

S5

S6

S7

55–58

60–62

54–57

54–57

55–58

58–60

60–62

Surface elevation (m amsl)

Table 2 Stratigraphic sequences within 30 m depth of the arsenic-contaminated area

2.0–4.0

10–18

7–15

5–7

2.0–3.0

2.0–4.0

0.5–1.0

Thickness of mud cover (m)

f1, f2, f3, f4, m1, m2, c1 and c2

Limited extent, adjoining parts of levees, representing proximal parts of flood plains

Part of arcuate shaped depressions filled with disturbed and redistributed sediments

Arcuate shaped wide lowlying areas and depressions. Typically cultivated paddy fields

f3, f4, m1, m2, c1 and c2

f3, f4, m1, m2, c1 and c2 and redistributed

Narrow curvilinear tracts representing palaeo-crossbar channels. Widely cultivated

Narrow curvilinear tracts representing palaeo-crossbar channels. Widely cultivated

Vast tracts adjoining the ridges. Sites of settlements, orchards, gardens and vegetable crop fields

Widespread as curvilinear ridges. Sites for settlements

Distribution in study area

f3, f4, m1, m2, c1 and c2

f3, f4, m1, m2, c1 and c2

f3, f4, m1 and m2

m1 and m2

Associated sediment facies in the mud cover overlying the SAS

Proximal flood plain sequence

Palaeochannel cut-off sequence with reworked flood plain mud/clay as the anthropogenic fill at the top

Palaeochannel cut-off sequence

Repeated palaeo-cross-bar channel sequence

Palaeo-cross-bar channel sequence

Braid bars, point bar plat form sequence, levee sequence

Scroll bars and point bar sequence

Depositional environment

Environ Earth Sci

123

m1,m2 c1,c2 f 3,f 4

m1,m2 c1,c2 f 3,f 4

7-15m

f 3,f 4 c1,c2

f 3,f 4 c1,c2 c1,c2

c1,c2

S2

S3

S5

S4 Anthropogenic fill

Shallow Aquifer Sand (f 1,f 2)

~ 30 m

c1,c2 f 3,f 4 c1,c2

S1

f 1,f 2,f 3,f 4, m1,m2, c1,c2

2-4 m

5-7 m

2-4 m

f 3,f 4

c1,c2 f 3,f 4 m1,m2

2-3 m

m1,m2 f 3,f 4 c1,c2

m1,m2

m1, m2

2-3 m

0.51.0 m

0.5-2 m

Environ Earth Sci

S6

S7

Alternation of sand (f 1-f 2)/silt(f 3-f 4) /mud(m1-m2)/clay(c1-c2)

Fig. 2 Seven stratigraphic sequences (S1–S7) identified in the area within 30 m below ground. f1–f4, m1–m2 and c1–c2 represent the sedimentary facies (for details see Table 1)

Dissolved arsenic in hand pump (< 30 m depth) < BDL - 10 µm/ L

Ganga River

50- 100 µm/ L

10 - 50 µm/ L

> 100 µm/ L Suhiya Suhiya Boundary between Active and Older Flood Plain

Bariswan Bariswan

Semaria Semaria Ojhapatti Ojhapatti

Dudhghat Dudhghat Boundary between Newer and Older Alluvium

Minor tributaries of Ganga River flowing through its flood plain Abandoned/ palaeochannel cut offs of the Ganga River

Rampur Rampur Nargada Nargada

Belauti Belauti Teghra Teghra Amrahi Amrahi Nawada Nawada

0 Older Alluvium

2.5

5 km

Kilometers

Fig. 3 Groundwater sampling locations and abundance of arsenic in the study area. The arcuate features with darker shade represent higher soil moisture along the nearly filled abandoned/palaeochannel cut-offs of the Ganga. The cross-bar channels, also in dark shades, frequent the OFP. Rain-water accumulates in these geomorphic features in the floodplain for considerable period of a year. The AFP represented by darker shade along the bank of the Ganga is frequently

flooded. The lighter shade areas are levees and point bars with higher elevation and thinner mud cover. These are the areas where most of the habitations are located. The groundwater samples collected from tube wells in the villages such as Semaria-Ojhapatti, Nargada, Dudhghat, Rampur and Teghra represent channel fill facies (strat. Seq. S6) with anthropogenic fill at the top

In the 1st category (n = 85), 94 % of the hand pumps located over the stratigraphic sequence S1 comprising the scroll bar ridges in the inner parts of the point bars with

SAS occurring at shallow depths (Fig. 4) show arsenic in the range of BDL-50 lg/L. Such lower concentrations are also observed at 87 % of the hand pumps (n = 71) located

123

Environ Earth Sci 70 (n=85)

(n=71)

(n=73)

(n=56)

% of sample distribution

60

50

where over the decades, the ground surfaces have been raised to the tune of 2–3 m by artificial filling (strat. Seq. S6). Such areas come under the 3rd category, forming the groundwater arsenic hotspots with *80 % of samples showing concentration beyond 50 lg/L. In few cases like those from Semaria-Ojhapatti village, the concentrations often exceeded 500 lg/L (max. 987 lg/L).

40

30

Discussion

20

The study area forms a part of SGP, a geomorphic unit denoting the alluvial plains between the Ganga River at the north and the Precambrian Highlands at the south. In Bhojpur and Buxar districts of the Bihar State, where the present study area is located, the Ganga River has shifted 5–20 km north/northwestward during the Holocene (Sahu et al. 2010). The channel entrenchment and shifting of the river have created a wide space for deposition of the Newer Alluvium. The width of the AFP varies from 2 to 15 km, being wider at west and narrowing down eastward. The OFP, dotted with meander cut-offs and abandoned channels, is richly cultivated for paddy. Arsenic concentration in groundwater in the area frequently exceeds the state regulating limit of 50 lg/L. Groundwater from both the floodplains (AFP and OFP) is found to be affected by arsenic contamination, contradicting the earlier observations of wide-scale contamination in the AFP (T0-Surface) and only of local-scale contamination in the OFP (T1-Surface) (Shah 2008). The OFP is densely inhabited in comparison to the AFP. Out of the total number of tested hand pumps, 32 % in AFP (n = 34) and 34 % in OFP (n = 249) have been found with arsenic concentration beyond 50 lg/L. The micro-depositional environments such as channel cut-off lakes, point bar platforms, point bars with ridges and depressions, and subdued natural levees have resulted in ten sedimentary facies. Spatial variation in the distribution of sedimentary facies has given rise to seven shallow alluvial stratigraphic sequences within the top 30 m of the sedimentary profile, which bears the aquifers with elevated groundwater arsenic concentrations. Clean grey sand (facies f1 and f2), representing the base of the stratigraphic sequences, forms a potential aquifer (Fig. 2), which is being extensively tapped through hand pumps and tube wells for rural drinking and irrigation. Groundwater arsenic contamination has got a specific relation with the Holocene geomorphic sub-features and the shallow stratigraphic sequences in the area. The contaminated areas coincide with the palaeochannel cut-offs of the Ganga and the narrower belt on the inner sides of the cut-offs, which were earlier occupied by cross-bar channels of various dimensions. The black clay/mud deposits (facies

10

Strat. sequence Category

S2 & S7

S1 1- Low arsenic

S3 & S4 2- Moderate arsenic

>100

50-100

BDL-10

10-50

>100

50-100

BDL-10

10-50

>100

50-100

BDL-10

10-50

>100

50-100

BDL-10

10-50

0 Groundwater 'As' conc. in µg/L

S6 3- High arsenic

Fig. 4 Relationship between frequency of water samples and abundance of arsenic in different stratigraphic sequences representing different geomorphic units in the floodplain. ‘n’ shows the number of groundwater samples tested

on palaeo-levee, point bar platform and proximal floodplain sequences (strat. seq. S2 and S7), where the mud cover overlying the SAS is predominantly made up of mud, silt and very fine sand (facies f3, f4, m1 and m2). The levee adjoining the channel cut-off lake in AFP (strat. seq. S2), however, shows comparatively higher contamination, where 31 % of tested sources (n = 16) exhibit arsenic levels [50 lg/L (max. 149 lg/L). Leaving this unit aside, the number of sources within 50 lg/L in the sequences S2 and S7 goes up to 93 % (n = 55). With the exception of levees along the active channel cut-off lakes, an average of *39 % of the samples in the environments as described above in category 1 (strat. seq. S1, S2 and S7) exhibit even \10 lg/L of arsenic in groundwater (Fig. 4). In the 2nd (n = 73) and 3rd (n = 56) categories, elevated concentrations of arsenic ([50 lg/L) occur in clusters, showing good correlation with the palaeochannels filled with mud/clay (strat. seq S3, S4 and S6). The villages located over the filled up cross-bar channels, representing the sequences S3 and S4, and particularly those adjacent to the clay-plugged palaeochannels (strat. seq S5 and S6), are significantly contaminated (Category 2). About 51 % of the sources have exceeded 50 lg/L of arsenic content and in 30 % of them, the arsenic concentrations have even exceeded 100 lg/L (max. 342 lg/L). Some villages such as Semaria-Ojhapatti, Dudhghat, Nargada, Rampur and Teghra are located on the clay-plugged palaeochannels,

123

Environ Earth Sci

123

2

(a) Thickness of top clayey sediments at surface (m)

R = 0.48 4.0 3.0 2.0 1.0 0.0 0

Thickness of top clayey sediments at surface (m)

(b)

30

(c)

60

2

R = 0.42

20.0 16.0 12.0

Known thickness data

8.0

Extrapolated data

4.0 0

Arsenic concentration (µg/L)

c1 and c2) reveal variable thickness from \1 m in crossbar channels to 10–15 m in palaeochannel cut-offs. The facies has higher organic carbon content (range 5.9–6.6 g/kg) than (range 0.8–6.1 g/kg) the brown floodplain mud (facies m1 and m2) and other sedimentary facies (range 0.1–1.2 g/kg, Table 1). Organic carbon increases structural stability of soil and its resistance to rainfall impact, which otherwise clogs the pores by particle breaking. The retaining of pores in the soil enhances the infiltration rate through it (Roose and Barthes 2001). Analyses show 8.5–12.2 Tritium Units in groundwater from SAS within 30 m depth, suggesting modern recharge (\40 years) (Saha et al. 2011b). A significant part of the recharge is possibly contributed from impounded surface water bodies such as channel cut-off lakes and other floodplain depressions as indicated by d18O distribution (Saha et al. 2011b). The surface water while percolating downward collects the organic carbon through dissolution from the clay bodies. The organic carbon might also be coming to the groundwater continuously at the interface between the aquifer and clay bodies in the palaeochannels. A correlation has been tried to obtain between the thickness of top clayey sediments and the arsenic concentration in groundwater in the shallow aquifer (Fig. 5a, b). The possible number of thickness data of such clayey layers has been obtained from available borehole lithologs, study of excavation/trench faces, field queries from local drillers and few extrapolated ones on the basis of surface morphology and shallow sedimentary facies variation. The plot reveals marginal increase in arsenic concentration in areas where the top clayey sediment is \4.5 m thick (Fig. 5a). However, as the thickness exceeds, rapid increase in arsenic concentration (often [80 lg/L) is noticed (Fig. 5b). Thick clayey unit at top of the sequence often represents channel fill deposits (facies c1, c2 and m1, m2), which also act as sink for organic carbon. In the scatter plots (Fig. 5a, b), the thicknesses of clayey sediments have been considered irrespective of the geomorphic units they belong to, which has led to moderate correlation coefficients (R2 = 0.48 and 0.42). The arsenic concentrations in hand pumps may vary even though the thicknesses of clayey tops remain closely similar. It reflects the availability/non-availability of organic carbon in the top argillaceous sediment. The assessment reveals higher correlation coefficient when it is made for a specific geomorphic unit. As for example, in Semaria-Ojhapatti village, where the hand pumps tap SAS below the thick clayey sediments (16–18 m) deposited in fluvio-lacustrine set-up, the plot between groundwater arsenic and thickness of top clayey sediments exhibits a better correlation (R2 = 0.82, Fig. 5c). The screens of the hand pumps tap SAS in all the seven stratigraphic sequences. The relation of the arsenic

600

1200

Arsenic concentration in groundwater ( µg/L) 1200

2

R = 0.82 1000 800 600 400 200 0 25

30

35

40

45

Depth of hand pump (m amsl)

Fig. 5 a The plot of thickness of top clayey sediments at surface overlying the SAS vs. groundwater arsenic concentration. A marginal increase in the dissolved arsenic takes place as the clayey sediment thickness increases up to 4.5 m. b Groundwater arsenic concentration, where the SAS is overlain by thicker clay at top. It reveals rapid increase in the level of dissolved arsenic in groundwater at clayey top thicknesses beyond 4.5 m. c A plot of depth of hand pump vs. groundwater arsenic concentration at Semaria-Ojhapatti village in Bhojpur district (for details see text)

concentration with different geomorphic features and the shallow alluvial stratigraphy has been depicted schematically in Fig. 6. Most of the hand pumps (range 87–94 %) in the stratigraphic sequences S1, S2 and S7, which cover *70 % of the area under meander scars, recorded lower arsenic concentrations (\50 lg/L). Exceptions do occur in the sequence S2 (levee) at the channel cut-offs forming active lakes. Moderate to high concentrations of arsenic (range 50–342 lg/L) have been noted at 51 % of the tested sources from the sequences S3 and S4 representing

Environ Earth Sci

Fig. 6 A schematic section (south to north) showing generalised topography, floodplain morphology and shallow alluvial stratigraphy. It depicts the concentrations of arsenic in hand pumps (within *30 m bgl) and their distribution in different stratigraphic sequences in the settlement areas. High arsenic (50–342 lg/L) has got an association with the hand pumps (depth 10–30 m bgl) in the villages settled on

anthropogenic fills over the channel fill sequence (strat. Seq. S6). The sandier areas in the scroll bar ridges, levees (particularly in OFP) and proximal floodplain sequences (strat. seq. S1, S2 and S7) often exhibit lower concentrations of arsenic (\50 lg/L). It is noteworthy that these environments together make *70 % of the area of a typical meander scar of the Ganga

abandoned cross-bar channels on the point bars. The channel fill sequence S6 with the anthropogenic fill at the top shows high concentrations of groundwater arsenic (range 50–987 lg/L). The arsenic concentration in groundwater has therefore been correlated with the availability of organic carbon concentration per unit width along longitudinal section of clayey sediment caps overlying the SAS (Table 2). Though the floodplain mud (facies m1 and m2) possesses moderate to high organic carbon, availability of total organic carbon is low due to their limited thickness. In case of the sequences S5 and S6 possessing thick black mud/clay (with moderate to high organic carbon), the total organic carbon is significant. Likewise, the sequences S3 and S4 are moderate in their organic carbon concentrations. The channel lakes of recent meander cut-offs are areas of active biomass accumulation (Fig. 6). In the settlement clusters on the banks (strat. seq. S2) of such lakes, localised water level trough is formed due to groundwater extraction. A part of the organic carbon of lake water and underlying sediment percolates downward and moves village-ward along with the groundwater under the influence of modified hydraulic gradient. The consumption of organic carbon by bacteria and the metabolism at the expense of dissolved oxygen in groundwater promote reductive dissolution of hydrated iron oxide (HFO) coatings on sediment grains, thereby releasing arsenic in groundwater. The reductive dissolution of HFO in arsenic-affected areas of MGP is

evident by high concentration of dissolved iron in groundwater (Shah 2008, 2013). Coexistence of redoxsensitive solutes such as As-III (arsenite, the highly mobile and poisonous form of arsenic), Fe-II, NH3 and elevated HS- in groundwater in MGP also indicates a reducing, post-oxic condition (Mukherjee et al. 2012). In arseniccontaminated areas, Kumar et al. (2010) have reported the mean ratio of As III/Astotal as 0.67, whereas Mukherjee et al. (2012) worked out a higher mean ratio of 0.90. The redox potential (Eh) values largely varying between -134 and 255 mV have been reported from the MGP (Kumar et al. 2010; Mukherjee et al. 2012), where the values between -200 and 100 mV are considered as moderate to highly reducing (Ferguson and Gavis 1972; Masscheleyn et al. 1991). The spread of the reducing environment depends upon the volume of organic carbon release, hydraulic conductivity of aquifer formation, groundwater flow direction, and the volume of fresh oxic water recharge that reaches the spreading anoxic front. The sandier areas along the scroll bar ridges and levees yield low concentration of arsenic as the oxic groundwater in such areas favours stability of hydrated iron oxide. There may be a source–distance relation between the arsenic concentration and the organic carbon-rich clay plugs in the stratigraphic sequence. Though the study broadly implies a decrease in groundwater arsenic concentration away from such clay bodies, it is difficult to ascertain conclusively at this stage

123

Environ Earth Sci

with the present set of data, since the studied floodplain is highly complicated with rapid spatial facies variation. Further detailed research is required to confirm the relation. Villages on artificial fills The villages are generally located on the sandier scroll bar ridges of the meander scars. Few villages such as SemariaOjhapatti, Dudhghat, Nargada, Rampur and Teghra are settled on 2–3 m thick artificial fills over palaeochannels of Ganga. The anthropogenic fills form a part of the stratigraphic sequence S6 at its top. The sequence is clayey up to 10–15 m depth (channel plug facies) and is marked with organic carbon-rich (5.9–6.6 g/kg) black mud. It is encountered mostly in OFP and groundwater arsenic concentration frequently exceeds 50 lg/L in areas occupied by such facies. Out of 56 samples collected from these villages, 55 samples exceeded the WHO standard of 10 lg/L, while 45 exhibited arsenic beyond the state regulating limit of [50 lg/L. Noteworthy fact is that 37 samples of the villages registered even [100 lg/L of arsenic in groundwater (max. 987 lg/L), particularly at Semaria-Ojhapatti village, where 92 % of the samples (n = 15) contain [100 lg/L of arsenic. The anthropogenic fill over S6 on which the village is located might be playing a role in enhancing the arsenic concentration in groundwater. The artificial fill might be creating armour of hardened surface, which facilitates more run off and allows less water to infiltrate downward. It encourages reducing environment below the settlement areas, conducive for arsenic release to groundwater. The decreased recharge from rainfall and induced groundwater flow towards such villages because of increasing development might be worsening the arsenic contamination scenario. The Bariswan village located on the cross-bar channels (strat.seq. S3 and S4) in the vicinity of channel plug sequence (strat.seq. S5 and S6) also reveals elevated levels of arsenic in groundwater.

Elevated concentrations ([50 lg/L) are observed close to the organic matter rich clayey deposits in the abandoned/ palaeochannel cut-offs, which are either filled or in the process of filling under lacustrine environments. The hotspots of arsenic contamination (max. 987 lg/L) have been found (in *80 % of the tested sources) in settlements located on anthropogenic fills over the clay/mud plugged palaeochannels (strat. seq. S6), often located in the OFP. Moderate to high levels of arsenic concentrations (max. 342 lg/L) are encountered (in 51 % of tested sources) in areas close to such palaeochannels (strat. seq. S5 and S6) and the areas traversed by cross-bar channels (strat. seq. S3 and S4). The prominent ridges on point bars, levees and point bar platforms (strat. seq. S1, S2 and S7), predominated by sand facies, are by and large low in arsenic. Most of the hand pumps (89–94 %) from these units exhibit arsenic concentrations \50 lg/L. The thickness of the dark grey to black coloured clay/ mud cover overlying the SAS and the availability of organic carbon are the controlling factors for release of arsenic in the shallow aquifers in MGP. The spatial variability of groundwater arsenic is closely associated with the variability of organic matter content in argillaceous sedimentary bodies (clay/mud facies) within the Newer Alluvium and their release to groundwater. The geomorphic and stratigraphic approach attempted in this study can be adopted in other areas of MGP for microlevel delineation of the areas vulnerable to groundwater arsenic contamination. Alternately saying, the methodology can be adopted to locate the arsenic-free shallow groundwater zones within the contaminated area. Acknowledgments The authors express sincere thanks to S. Gupta and R. C. Jain for their help. Thanks are also extended to V. Srivastava and D. S. Mishra for their suggestions. The opinions expressed in this paper are of authors of their own. The research has been carried out as a part of the PhD thesis work of the first author at Banaras Hindu University, India.

Conclusion and recommendations

References

The physical dynamics of the Ganga River in MGP, such as its meandering and migration during the Holocene have formed a wide fluvio-lacustrine landscape conducive for release of arsenic from solid phase to groundwater. The distribution pattern of groundwater arsenic bears significant correlation with the floodplain geomorphologic elements. In contrary to the earlier observation of wide-scale contamination in the AFP (T0-Surface) alone, the present study reveals that the contaminated sources are distributed in both the AFP (T0-Surface) and OFP (T1-Surface), where the shallow contaminated aquifers are made up of Himalayan derived grey to dark grey coloured finer sediments.

Acharya SK (2004) Arsenic trends in groundwater from quaternary alluvium in Ganga plain and Bengal basin, Indian subcontinent: insights into influences of stratigraphy. Gondwana Res 8:55–66 Acharya SK, Shah BA (2007) Arsenic- contaminated groundwater from parts of Damodar fan- delta and west of Bhagirathi River, West Bengal, India: influence of fluvial geomorphology and Quaternary morphostratigraphy. Environ Geol 52:489–501 Allen JRL (1965) A review of the origin and characterisation of recent alluvium sediments-. Sedimentology 5:89–191 Berg M, Tran HC, Nguyen TC, Pham HV, Schertenleib R, Giger W (2001) Arsenic contamination of ground and drinking water in Vietnam: a human health threat. Environ Sci Technol 35:2621–2626. doi:10.1021/es010027y BGS, DPHE (2001) Arsenic contamination of groundwater in Bangladesh. In: Kinnibugh DG, Smedley PL (eds) BGS

123

Environ Earth Sci Technical Report WC/00/19 British Geological Survey, Keyworth, 267 Burgess WG, Hoque MA, Michael HA, Voss CI, Breit GN, Ahmed KM (2010) Vulnerability of deep groundwater in the Bengal Aquifer system to contamination by arsenic. Nature Geosci 3:83–87 CGWB, PHED (2005) A report on status of arsenic contamination in groundwater in the state of Bihar and action plan to mitigate it Govt of India and Public Health Engineering Department, Govt of Bihar Central Ground Water Board Chakraborty C, Chattopadhyay GS (2001) Quaternary geology of South Ganga Plain in Bihar. Indian Min 55(3 and 4):133–142 Chakraborty D, Mukherjee SC, Pati S, Sengupta MK, Rahman MM, Chowdhury UK, Lodh D, Chanda CR, Chakraborti AK, Basu GK (2003) Arsenic groundwater contamination in Middle Ganga Plain, Bihar, India: a future danger? Env Health Perspective 111:1194–1201 Dayal N (1997) Quaternary geology and geomorphology of IndoGangetic Plain, Bihar. Golden Jubilee Vol., Dept. Geol., Patna Univ., Patna, India, pp 29–33 Ferguson JF, Gavis J (1972) Review of the arsenic cycle in natural waters. Water Res 6:1259–1274 Gigera W, Berga M, Phamb HV, Duongb HA, Tranb HC, Caob TH, Schertenleiba R (2003) Environmental analytical research in Northern Vietnam––A Swiss Vietnamese cooperation focusing on arsenic and organic contaminations in aquatic environments and drinking water. Chimia 57:529–536 Govt. Of Bihar (2007) State of Environment Report, Bihar by Bihar State Pollution Control Board, Patna and Department of Environment and Forest Hoque MA, Khan AA, Shamsudduha M, Hossain MS, Islam T, Chowdhury SH (2009) Near surface lithology and spatial variation of arsenic in the shallow groundwater: southeastern Bangladesh. Environ Geol 56:1687–1695 Hudson PF, Heitmuller FT (2003) Local- and watershed-scale controls on the spatial variability of natural levee deposits in a large.ne-grained floodplain: lower Pa´nuco basin Mexico. Geomorphology 56:255–269 Johnston RB, Hanchett S, Hoque Khan M (2010) The socio-economic of arsenic removal. Nature Geosci 3:2–3 Khan AA, Hoque MA (2002) Quaternary Paleo-geography and Geohazard Scenario of the Bengal Delta of Bangladesh. Oriental Geographer 46:1–16 Kumar P, Kumar M, Ramanathan AN, Tsujimura M (2010) Tracing the factors responsible for arsenic enrichment in groundwater of the Middle Gangetic Plain, India, a source identification perspective. Environ Geochem Health 32:139–146 Malik S, Biswas B (2014) Geomorphology, Quaternary morphostratigraphy and spatial distribution of arsenic contamination in groundwater, a case study of Katwa-II, Burdwan, West Bengal. J Environ Earth Sci vol 4, No.5 Masscheleyn PH, Delaune RD, Patrick WH Jr (1991) Effect of redox potential and pH on arsenic speciation and solubility in a contaminated soil. Environ Sci Technol 25:1414–1419 McArthur JM, Banerjee DM, Hudson-Edwards KA, Mishra R, Purohit R, Ravenscroft P, Cronin A, Howarth RJ, Chatterjee A, Talukder T, Lowry D, Houghton S, Chadha DK (2004) Natural organic matter in sedimentary basins and its relation to arsenic in anoxic groundwater: the example of West Bengal and its worldwide implications. Appl Geochem. 19:1255–1293. doi:10.1016/j.apgeochem.2004.02.001 McArthur JM, Sikdar PK, Hoque MA, Ghosal U (2012) Wastewater impact on groundwater: Cl/Br ratio and implications for arsenic pollution in the Bengal Basin and Red River Basin Vietnam. Sci Total Environ 437:390–402

Mukherjee A, Scanlon BR, Fryar AE, Saha D, Ghosh A, Chowdhury S, Mishra R (2012) Solute chemistry and arsenic fate in aquifers between the Himalayan foothills and Indian craton (including central Gangetic Plain); influence of geology and geomorphology. Geochem et cosmochimica acta, 90-283–302 Nayak B, Das B, Mukherjee SC, Pal A, Ahmed S, Hossain NA, Meity P, Dutta RN, Dutta S, Chakraborti D (2008) Groundwater arsenic contamination in the Sahebganj district, Jharkhand state India in the Middle Ganga Plain and adverse health effects. Toxicol Environ Chem, doi:10.1080/02772240701655486 Neumann RB, Ashfaque KN, Badruzzaman ABM, Ali MA, Shoemaker JK, Harvey CF (2010) Anthropogenic influences on groundwater arsenic concentrations in Bangladesh. Nature Geosci 3:46–52 Nickson RT, McArthur J, Burgess W, Ahmed KM, Ravenscroft P, Rahman M (1998) Arsenic poisoning of groundwater in Bangladesh. Nature 395:338 Polya DA, Gault AG, Diebe N, Feldmann P, Rosenboom JW, Gilligan E, Fredericks D, Milton AH, Sampson M, Rowland HAL, Lythgoe PR, Jones JC, Middleton C, Cooke DA (2005) Arsenic hazard in shallow Cambodian groundwater. Miner Mag 69:807–823 Prakash Om, Sinha AP, Verma NP, Reddy BSS (1990) Quaternary geological and geomorphological mapping of the Ganga-Sone alluvial belt in Aurangabad, Bhojpur, Jehanabad, Patna and Rohtas districts Bihar. Rec Geol Surv India 123(3):8 Raju NJ (2012) Arsenic exposure through groundwater in the Middle Ganga Plain in the Varanasi environs, India: a future threat. J Geol Soc India 79:302–314 Ravenscroft P, Burgess WG, Ahmed KM, Burren M, Perrin J (2005) Arsenic in groundwater of the Bengal Basin, Bangladesh: distribution, field relations, and hydrogeological setting. Hydrogeol J 13:727–751 Roose E, Barthes B (2001) Organic matter management for soil conservation and productivity restoration in Africa: a contribution from francophone research, 61, 159–170 Saha D (2009) Arsenic groundwater contamination in parts of Middle Ganga Plain Bihar. Curr Sci 97(6):753–755 Saha D, Dhar YR, Sikdar PK (2008) Geochemical evolution of groundwater in the Pleistocene aquifers of the South Ganga Plain Bihar. J Geol Soc Ind 71:473–482 Sahu S (2013) Hydrogeological conditions and geogenic pollution in parts of western Bihar. PhD thesis submitted to Department of Geology, Banaras Hindu University, Varanasi. Uttar Pradesh Shah BA (2008) Role of quaternary stratigraphy on arsenic contaminated groundwater from parts of Middle Ganga Plain, UP–Bihar India. Environ Geol 53:1553–1561 Saha D, Dhar YR, Vittala SS (2009) Delineation of groundwater development potential zones in parts of Marginal Ganga Alluvial Plain in South Bihar, Eastern India. Environ Monit Assess 165:179–191 Saha D, Dwivedi SN, Singh RK (2014) Aquifer system response to intensive pumping in urban areas of the Gangetic plains, India: the case study of Patna. Environ Earth Sci 71:1721–1735. doi:10. 1007/s12665-013-2577-7 Sahu S, Dwivedi SN (2012) Exploring possible alternative safe sources of water for irrigation in arsenic affected areas in Middle Ganga Plain, India. Abstract Volume. World Water Week in Stockholm. 26–31, pp 172–173 Saha D, Sahu S, Chandra PC (2011a) Delineating arsenic-safe deeper aquifers and their hydraulic parameters in parts of Middle Ganga Plain, Eastern India. Environ Monit Assess. doi:10.1007/s10661010-1535-z Sahu S, Saha D (2013) Geomorphologic, stratigraphic and sedimentologic evidences of tectonic activity in Sone-Ganga alluvial

123

Environ Earth Sci tract in Middle Ganga Plain, India. J Earth System Sci (Waiting for publication, JESS-D-13-00365) Saha D, Shukla RR (2013) Genesis of arsenic rich groundwater and the search for alternative safe aquifers in the Gangetic Plain India. Water Environ Res 85(12):2254–2264 Singh RL, Singh KN (1991) Middle Ganga Plain in Singh in India––a regional geography. Singh RL (Ed) National Geographical Society of India. Ram Printograph (India) C-114 Okhla Industrial Area, Phase I, New Delhi-110020. pp 183–251 Sengupta S, Mukherjee PK, Pal T, Shome S (2004) Nature and origin of arsenic carriers in shallow aquifer sediments of Bengal Delta, India. Environ Geol 45:1071–1081. doi:10.1007/s00254-0040965-8 Sahu S, Raju NJ, Saha D (2010) Active tectonics and geomorphology in the Sone-Ganga alluvial tract in mid-Ganga Basin India. Quat Intern 227:116–126 Saha D, Sinha UK, Dwivedi SN (2011b) Characterization of recharge processes in shallow and deeper aquifers using isotopic signatures and geochemical behavior of groundwater in an arsenicenriched part of the Ganga Plain. Appl Geochem 26(4):432–443 Saha D, Sreehari SMS, Dwivedi, SN, Bhartariya KG (2010) Evaluation of hydrogeochemical processes in arsenic contaminated alluvial aquifers in parts of Mid-Ganga Basin, Bihar, Eastern India. Environ Earth Sci. doi:10.1007/s12665-009-0392-y

123

Sharma AK, Tjell JC, Mosbæk H (2006) Health effects from arsenic in groundwater of the Bengal delta: effects of iron and water storage practices. J Environ Geosci 13(1):17–29 Saha D, Upadhyay S, Dhar YR, Singh R (2007) The aquifer system and evaluation of its hydraulic parameters in parts of South Ganga Plain, Bihar, India. J Geol Soc India 69:1031–1041 Van Geen A, Zheng Y, Versteeg R, Stute M, Horneman A, Dhar R, Steckler M, Gelman A, Small C, Ahsan H, Graziano J, Hussein I, Ahmed KM (2003) Spatial variability of arsenic in 6,000 tube wells in a 25 km2 area of Bangladesh. Water Resour Res 39:1140–1155. doi:10.1029/2002WR001617 Von Bro¨mssen M, Jakariya M, Bhattacharya P, Ahmed KM, Hasan MA, Sracek O, Jonsson L, Lundell L, Jacks G (2007) Targeting low-arsenic aquifers in Matlab Upazila Southeastern Bangladesh. Sci Tot Env 379:121–132 Weinman B, Goodbred SL Jr, Zheng Y, Aziz Z, Steckler M, van Geen A, Singhvi AK, Nagar YC (2008) Contributions of flood plain stratigraphy and evolution to the spatial patterns of groundwater arsenic in Araihazar Bangladesh. Geol Soc Am Bull 120:1567–1580 WHO (1993) Guidelines for drinking water quality, 2nd edn. v 1, WHO, Geneva