Arsenic in Ground Water in Parts of Middle Ganga Plain in Bihar-An Appraisal

INTRODUCTION

Ground water in different parts of the world has now been identified with unacceptablyhigh concentration of arsenic, known for its carcinogenic health effects. The areascontaminated with arsenic in groundwater include Bangladesh and parts of India, China(including Taiwan), Myanmar, Nepal, Vietnam, Cambodia, Argentina, Chile, Mexico, SouthWest United States, Hungary and Romania (Chatterjee et al 2005). In India, Bangladeshand some other developing countries, regulatory limit of arsenic in potable water is beingconsidered as 0.05 mg/L. However, WHO (1993) has recommended a drinking waterguideline value of 0.01 mg/L, which has also been endorsed by Bureau of Indian Standards(BIS 2003).

Arsenic groundwater contamination in Bihar was first detected in the year 2002.Subsequent investigations in the Gangetic Plain of Bihar revealed its wide occurrence,affecting 57 blocks in 15 districts of the state. The patches of high groundwater arsenic(>0.05 mg/L) zones are confined in Newer Alluvial belt along the river Ganga affecting boththe active and the older flood plains. The contamination is confined within the top 50 m ofthe thick multi-cyclic sand, clay, sandy clay and silty clay sequence, jeopardizing the handpump based rural drinking supply.

SPATIAL GROUNDWATER ARSENIC DISTRIBUTION

In Ganga Plain, in the upstream of Rajmahal Hills, groundwater arsenic exceeding0.05mg/L was initially detected in 2002, from two villages, Semariya- Ojhapatti and Doodhghat, in Bhojpur district of Bihar, on the southern bank of the River Ganga(Chakraborty et al 2003). The villages are located on a monotonously flat, flood-prone tractof the Son-Ganga interfluve region. Investigations by CGWB, immediately after reporting of this, indicated maximum concentration as 0.178 mg/L. All the affected wells are hand pumps within a depth range of 20-35 m bgl. The dug wells (depth
The Hydrogeological studies in Bhojpur district has revealed wide variation in concentration, resulting in patchyness in distribution. The hotspots (As >0.05 mg/L) are confined within the Newer Alluvium (CGWB 2006). With an understanding that the North Bihar belt adjoining the arsenic contaminated Tarai belt of Nepal (Tandukar et al 2001) is also affected, a detailed analyses has been carried out by UNICEF under technical guidance from CGWB. In nine districts of north Bihar bordering Nepal, the analysis (total sample ~3100) has revealed arsenic load below 0.05 mg/L. Subsequently a blanket survey of the spot sources in a 20 km wide corridor along the course of the river Ganga has been carried out by PHED, Govt. of Bihar. Sampling from the reported hotspots, analyses and hydrogeological investigations renders an elaborate understanding on the spatial distribution of arsenic hotspots (CGWB and PHED 2005). Out of ~82000 samples analysed, 11% exceed 0.05 mg/L. Fifty seven blocks, in 15 districts, located on both the banks of Ganga are affected (Table 1, Fig. 1).

In a district-wise assessment, Bhojpur and Buxar are found to be relatively more affected,where 16-20% of the samples exceed 0.05 mg/L. Other districts following are Bhagalpur,Katihar and Samastipur where 11-15% of the samples are contaminated. The lesseraffected districts are Saran, Begusarai and Vaishali where

GEOLOGICAL FRAMEWORK

About 90 % of the geographical area of the state is underlain by Gangetic alluvium laidduring the Quaternary Period and forms a major part of the Middle Ganga Plain (MGP). Inthe west the MGP enters in Uttar Pradesh and in the east it merges with the Lower GangaPlain/ Bengal Delta Plain in West Bengal and Bangladesh. The Precambrian Highlands inthe south forms the southern border whereas in the north it merges with the Tarai belt,geographically located in Nepal along the foothills of lesser Himalayas. The Quaternarydeposits in MGP having their provenance both in the Himalayas and the Peninsular craton,are conveniently divided into the Newer Alluvium (Holocene) and the Older Alluvium(Pleistocene) (Acharya 2005).

The Older Alluvium bordering the Precambrian Highlands in south is assigned UpperPleistocene to Lower Holocene age by Chakraborty and Chattopadhyay (2001) whereas itoccurs as discontinuous patches in the North Ganga Plain. The Younger (Newer) Alluviumis of Middle to Upper Holocene age and at many places the transition is marked with anerosional unconformity. Recent deposits are confined in the flood plains of the Ganga Riverand as wide patches in North Ganga Plain, in the flood-prone tract of the Ghaghra, Kosi,Gandak, Burhi Gandak and Baghmati Rivers which are perennial and carry substantialsediment load from the lesser and the upper Himalayas.

The thickness of the Quaternary sediments increases from a couple of meters along thefringe areas in south to > 700 m in the axial part of MGP and increases further towardsnorth. Drilling by CGWB confirms the thickness of Quaternary sediments as > 450 m atBegusarai. The deposits are of fluvio-lacustrine origin, are made up of multi-cyclic sand,clay, sandy clay, in several fining upward sequence. The Older Alluvial deposits areyellowish to yellowish-brown in colour in contrast to gray coloured Holocene sediments.

ARSENIC CONTAMINATION SCENARIO AND QUATERNARY MORPHO-STRATIGRAPHY

The major rivers in the Ganga Plain in Bihar state exhibit narrow river valleys which areincised on to upland terraces (T2) developed on the Older Alluvium (Acharya 2005). TheNewer Alluvial deposits along the river channels particularly the Ganga are incised asnarrow terraces (T1). The T1 surface generally lies above the active flood plain (T0), which is subject to flooding, overtopping of banks during monsoon. The T1 surface is referred asolder flood plain.

Arsenic in groundwater exceeding 0.05 mg/L is confined in the Newer Alluvial depositsalong the Course of the River Ganga. The Pleistocene deposits exposed both in the Southand the North Ganga Plains are low in groundwater arsenic load and nowhere theconcentration has been reported to exceed 0.05 mg/L. A study in Sone-Ganga interfluveregion covering Bhojpur district, maximum concentration in the Older Alluvium has beenfound as 0.007 mg/L (Saha et al 2009). About 59% (n=17) of the samples from OlderAlluvium reported BDL. In the contiguous Newer Alluvial belt 61% and 41% of the samples(n=60) has exceeded 0.01 and 0.05 mg/L.



Earlier workers (Acharya 2005) opined that potential risk areas are confined to the activeflood plains in the Newer Alluvium. Study by CGWB in parts of Bhojpur district reveals that arsenic exceeding 0.05 mg/L is affecting both the active and the older flood plains of the river Ganga. The pattern of arsenic distribution bears a sympathetic relationship with the Quaternary geomorphic and stratigraphic developments. Several sedimentary and stratigraphic facies have been identified and relation between arsenic groundwater contamination and shallow alluvial stratigraphy has been elaborated. The areas in and around the abandoned/cut-off palaeochannels of Ganga are the arsenic hotspots, as hasbeen observed in Bhojpur and Buxar districts. The cutoff palaeo-channels in theOlder Flood Plain are filled and possesses a stratigraphic facies comprising floodplain/lacustrine (channel deep) mud/clay deposits rich in organic carbon (Facies 1,Fig.2) and are sympathetic to arsenic contamination. However, some of them stillexist as curvilinear depressions, which hold water seasonally. Most of those in the Newer Flood Plain presently form enclosed water bodies in which sedimentation process is going on. High-arsenic groundwater has also been reported from the villages in and around these cut-off lakes. The settlements at Semaria Ojhapatti and Nargada that are within the span of cut-off palaeo-channels bear significantly high levels of groundwater arsenic. These settlements possess a stratigraphic facies comprising thick (15-18 m) flood plain mud/lacustrine deposits overlain by a reworked artificial fine sediment fill (Facies 2, Fig. 2).

Arsenic concentration shows wide spatial variation as also observed in the BDP(Bhattacharya et al 1997). In Semaria–Ojhapatti village in Bhojpur district the variation inconcentration from hand pumps (depth 20-35 m) has been observed as 54 times. In theeastern part at Gosaidaspur village in Bhagalpur district, located in the active flood plain ofthe Ganga, the concentration in hand pumps (20-30 m) varies from 1.8mg/L within a distance of 800 m (Fig 3).

DEPTH DISTRIBUTION OF ARSENIC IN GROUNDWATER

Arsenic concentration in groundwater reduces with depth (Saha et al., 2010). In the BDPhighest concentration and greatest spatial variability occurs in a few tens of meters belowthe ground surface and decrease rapidly below 100 m (Ravenscroft et al 2005). In the MGPin Bihar, because of availability of potential aquifers at shallow depth and shallow water level(generally

HYDROCHEMISTRY OF ARSENIC CONTAMINATED GROUNDWATER

Arsenic-contaminated groundwater is near-neutral to mildly acidic and dominated byalkaline earth (Ca²⁺ + Mg²⁺) and weak acid (HCO_3-). Hydrogeochemical characteristics of Newer Alluvium and those from low-arsenic Older Alluvial deposits in Bhojpur district have been studied (Saha et al 2009). Marginally higher load of HCO_3- (av. 295 mg/L) is reported from samples in the Newer Alluvium compared to that of the Older Alluvium (av. 263.8 mg/L) while SO4^2-+Cl- load is higher in the samples from Older Alluvium. The cation chemistry in Older Alluvium is marked by dominance of Na⁺ over Ca²⁺, whereas in Newer Alluvium equal prevalence of Ca²⁺ and Na+ exists (Table 2).


The groundwater in the contaminated belt are categorized based on Cl-, SO42-and HCO3- concentrations as normal chlorine (
Most frequently observed facies is Mg-HCO3, followed by Ca-HCO3 and Ca-Na-HCO3. TheCa2+ and Mg2+ dominated facies exhibit frequent incidence of high arsenic (> 0.05 mg/L).Four facies are particularly found to be arsenic affected Ca-HCO3 (BDL- 0.620 mg/L), Mg-HCO3 (BDL- 0.227 mg/L), Ca-Mg-HCO3 (BDL-0.550 mg/L) and Mg-Ca-HCO3 (BDL-0.270mg/L). The Older Alluvial areas are dominated by Na+ dominated facies, viz, Na-Mg-HCO3,Na-HCO3 and Na-Ca-HCO3.


Hydrogeochemical evolution of groundwater in arsenic contaminated Newer Alluvium parthas been elaborated by Principal Component Analysis (PCA). The analysis involved varimaxrotation to achieve rotated factor matrix. The rotation was used for better interpretation bymaximizing the difference between the variables (Lee et al 2001). Principal Component1 (PC1) accounts for 29.08% of the variance and is contributed mainly by EC, HCO3- and Mg2+, indicating infiltration of rainfall and seepage from surface water bodies (Table 3). High Mg2+ loading in PC1 indicate dissolution of ferromagnesian minerals and detrital dolomites. PC2 accounts for 20.01% of the variance and characterized by high positive loadings ofCa2+, As (total) and Fe (total), revealing possible same mobilization path of As and Fe.Incidentally As and Fe shows strong positive correlation (r2= 0.674)(Fig 6). Loading ofHCO3- (0.364) indicates its role for releasing As and Fe in the aqueous phase.

HYDROGEOLOGY OF ARSENIC CONTAMINATED AREA

Arsenic contaminated areas in the state lies in the upper part of high potential multi-groupaquifer system. Visual analyses of drill-cut samples from 22 deep bore wells (~300 m)drilled by CGWB and 16'/64' electrical and self potential logging has helped understandthe aquifer geometry. The litholog of Bharauli, an arsenic contaminated village in Bhojpurdistrict, reveals a sequence of Quaternary fluvio-lacustrine deposits down to 250 m depth(Fig.7). At the top, a 13 m thick soil and clay sequence is followed by sandy clay, whichgrades to fine to medium sand up to 69 m bgl. A thick medium to coarse sand sequenceunderlies and continues till the appearance of light yellow colored clay and sandy clay(cumulative thickness 25 m). This clay and sandy clay layers appear to have lesspermeability and together have been referred as middle clay .



A second sequence of fine to coarse sand appears below the middle clay and continues up to 250 m below ground, with occasional gravel beds confined at two depth zones; 183-209 m bgl and 230-238 m bgl. No significant clay bed appears within the second sand sequence, except thin clay partings (~10 cm thick) appeared in fine to medium sand layers, indicating temporary low energy conditions. Various litho-units form a two-tier aquifer system separated by the middle clay. The upper aquifer ranges from 0 to 106 m bgl and lower aquifer starts at 130 m bgl. This two-tier aquifer system prevails in the inter-stream region of the Sone and the Ganga, covering Bhojpur and Buxar districts. Hydrogeologic section between karnamipur to Maner reveals spatial continuity of the two-tier aquifer system separated by 15-33 m thick clay (Fig8). Aquifer configuration, however, is different in the northern part of the river Ganga to the east of Sone-Ganga interfluve. The litholog of Madudabad in Samastipur district reveals continuance of the upper aquifer upto 120 m bgl. Clay and sandy clay predominate the lithology at depth. Aquifers of 12-15 m thickness are found embedded within the thick argillaceous deposits (Fig.9)

Groundwater in the arsenic-contaminated shallow aquifers (within 50 m bgl) remains underunconfined condition. The hydrograph network station measurements reveal shallowhydraulic head indicating effluent nature of river Ganga in this part. During the premonsoonthe water level remains between 5.0 and 10.0 m bgl, except in central parts inregions adjoining Ganga-Gandak confluence and parts of Buxar and Bhagalpur districtswhere it remains between 2.0 and 5.0 m bgl. During the post-monsoon the level rangesfrom 2.0 to 5. 0 m bgl except in few parts of Buxar, Bhojpur, Samastipur, Patna, Mungerand Bhagalpur districts where it remains between 5.0 and 10.0 m bgl.

In the arsenic contaminated Newer Alluvium belt in Bhojpur district a detailed studyindicates water levels of shallow contaminated aquifers found to vary between 5.1- 6.32 m(av 5.31m) and 3.2 to 4.43 m (av 3.78 m) during the pre and post monsoon seasonsrespectively. Groundwater flows towards the Ganga River with a gradient of 6x10-4indicating sluggish movement. The head of the low-arsenic lower aquifer remains higherthan the water levels representing shallow aquifer, 4.01–6.57 m during pre-monsoon and2.56 -4.01 m during post-monsoon seasons.

Long-duration pumping tests have been carried out in 16 exploratory wells. In three wells(Maner, Bharauli and Karnamepur) in Bhojpur district tapping the lower aquifer longdurationpumping test data have been interpreted by curve-matching methods of Walton(1962) and Hantush (1955), considering aquifers as semi-confined. Transmissivity varyfrom 6009 to 6985 m²/day (Table 4). Hydraulic conductivity ranges between 64.88 to 82.04 m/day. Storage coefficient values (6.4x10^-4 to 2.5x10^-3) indicate semi-confined to confined nature of the aquifer.

The water level behavior and the aquifer configuration reveal unconfined mode ofgroundwater occurrence in arsenic-contaminated shallow aquifer. Hydraulic conductivity ofthe shallow zone has been determined based on the grain-size parameters using theestablished equations by Lambe (1958) and Breyer (1964) for Bharauli well. In Lambe’sequation d10 is an important input whereas in Breyer’s method Uc (uniformity coefficient)plays a significant role. Porosity is the most critical input in determining K by Lamb’smethod. Based on published values of porosity for different size-grades of sand (Morris andJohnson 1967; Driscoll 1986) values of 39% and 36% are considered suitable for mediumto coarse sand and for very coarse sand (with occasional gravel) respectively. The hydraulicconductivity ranges between 37 and 92 m/day.

RECHARGE MECHANISM AND AGE OF GROUNDWATER IN DIFFERENT AQUIFERS

From the aquifer configuration it appears that the recharge paths are different for theshallow contaminated aquifers and the deep low-arsenic aquifers. The recharge mechanismhas also a bearing on mobilization of arsenic in groundwater. Isotope analyses, both stable(¹⁸O/¹⁶O, ¹³C) and radioactive (¹⁴C, ³H) have been carried out for aquifer-specific samples from Bharauli, Semaria- Ojhapatti, Bariswan and Sinha Gram in the arsenic contaminated area of Bhojpur district (CGWB and BARC, 2009).

Tritium (³H) concentration of the shallow groundwater (generally 3.42-10.13 TU) reveals a substantial component of modern recharge and the age has been estimated as less than 40 years. Carbon-14 concentration ranges widely, between 29.97 and 164 pMC, showing a positive correlation with depth of groundwater (Fig10). The deeper low-arsenic aquifer has 14C concentration between 29.97 and 77.61 pMC. Based on 14C concentrations, the age of the groundwater from deeper low-arsenic aquifers have been worked out as ~3000 years. The older age supports less permeable nature of the middle clay, holding the deeper aquifers under semi-confined to confined condition.

The δ¹⁸O (‰ VSMOW) values vary from -7.69‰ to -5.52‰ whereas from shallow aquifers it remains in the range of -7.06‰ to -5.52‰. Three samples from lower aquifers exhibit a δ¹⁸O range of -7.69‰ to -7.28‰ (Fig11), with an average value of -7.53‰, indicating that they are significantly depleted in ¹⁸O than the upper aquifer. δ²H (‰ VSMOW) varies in shallow aquifer from - 46.96‰ to -36.52‰. Deeper aquifer samples represent still lower values (-47.95‰ to -43.14‰). The difference in δ¹⁸O and δ²H (‰ VSMOW) values betweenshallow and the deeper groundwater indicates difference in hydrostratigraphy andpossibly the climatic regime under which the recharge took place. The plots ofδ¹⁸O vs δ²H fall both above and below the Global Meteoric WaterLine (GMWL) of Craig (1961) (Fig.11). Minor deviation from the GMWL indicates some evaporation of rainfall, prior to or during infiltration, or there might be some mixing of the infiltrating water with the pre-existing soil-moisture that has experienced several cycle of evaporation (Allison, 1982). A part of the recharge is contributed from natural surface water bodies, community tanks as well as recirculation of groundwater withdrawn for irrigation, which appears to be enriched in heavier stable isotopes due to evaporation.

CONCLUSIONS

Arsenic contamination in Groundwater beyond the regulatory limit of 0.05 mg/L is posing achallenge to the water supply in the state of Bihar. The contamination is affecting theshallow aquifer, within ~50 m bgl, which are the life line of hand pump based ruraldrinking supply.

Elevated arsenic load, are confined in the flood plain, both in the active (flood prone) andthe older flood plain of the Ganga River. The Older Alluvium forming the Marginal PlainUpland Surface (Singh 2004) in South Ganga Plain and Dissected Upland InterfulveSurface in the North Ganga Plain have low arsenic load. In Bhojpur district maximumarsenic concentration in Older Alluvium has been detected as 0.007mg/L. However,localized high arsenic has also been detected in Darbhanga, Kishenganj far away from thepresent course of the Ganga.

Arsenic distribution is marked with wide spatial variability resulting in patchyness indistribution. No relation has been established between the concentration level of arsenicand groundwater flow direction. Depth wise there is a significant reduction in concentrationbeyond 40 m bgl. The aquifers at depth (> 80 m bgl), which appears to be of Pleistoceneage are contamination free even considering the WHO (1993) guidelines.

The drill-cut samples from the arsenic contaminated Holocene deposits are gray colouredand rich in organic matter, indicating reducing environment. Deposition of organic matteris facilitated by numerous cut-offs, abandoned channels and back-swamps which holdwater for a significant part of the post-monsoon period. Grain size distribution of sandsreveals a fluvio-lacustrine depositional environment during the Holocene period.Chemically the groundwater is mildly acidic and fresh with TDS generally remaining within800 mg/L. The anionic chemistry is dominated by HCO_3- while the cationic chemistry is marked by dominance of Ca²⁺ and Mg²⁺. Groundwater with high arsenic load is marked with Ca-HCO₃, Mg-HCO₃ and Ca-Mg-HCO₃ facies. PCA of the major chemical constituents in Bhojpur district reveals high loadings of As (total), Ca²⁺ and Fe (total), and significant HCO_3- loadings in PC2. Release of As and Fe in aqueous phase is related to same mobilization path, which is supported by positive correlation between them.

The arsenic contaminated Newer Alluvium belt exhibits flat topography with shallow waterlevels (3- and so produce the relationship between AS and HCO_3- in shallow groundwater The deeper aquifers hold old water (> 3000 years) as indicated by 14C activity and appear tohave its recharge from far off places. Presence of sub-regional scale clay bed in Sone-Ganga interfluve region in Bhojpur and Buxar districts inhibits direct vertical percolation of arsenic contaminated water from shallow level to the deeper aquifer system beyond 120 m bgl. In other affected areas of Middle Ganga Plain, the arsenic free deeper aquifers are of good potential, where T remains > 5000 m²/day. Groundwater in these aquifers remains under semi-confined to confined condition and can be developed for community scale water supply through deep tube wells.

RECOMMENDATIONS

he spatial variation of arsenic concentration in groundwater, both horizontal andlongitudinal, warrants detailed investigation. A clear understanding of the mobilizationpath of arsenic in groundwater in the Ganga Plain is yet to be established. The reasons forlesser degree of contamination in the flood plains of the other major rivers draining theMiddle Ganga Plain, like Kosi and Gandak Rivers, as compared to the Ganga stem floodplain, also needs to be established. Sub-basin scale sustainable yield of the low-arsenicdeeper aquifers for community-scale water supply is to be worked out. The rechargemechanisms in the contaminated shallow aquifers and low-arsenic deeper aquifers needselucidation assimilating isotopic techniques both in the North and the South Ganga Plain.Clinical manifestations of prolonged arsenic exposure through drinking sources are alsoreportedly not as severe in the Middle Ganga Plain, as has been documented in WestBengal and Bangladesh. Studies throwing light on the clinical aspects are also desiredalong with ascertaining the possibility of arsenic fixation in the food chain.

ACKNOWLEDGEMENT

The authors take this opportunity to express their sincere thanks to Chairman, Member(SAM) and Member (SML) CGWB for their kind support and inspiration. Thanks are due toDr P.C.Chandra (Regional Director) for his constructive guidance and valuable support incarrying out the work and his fruitful suggestions in preparing the manuscript. Sincerethanks are extended to R.S.Singh (Ex-Regional Director) for taking keen interest in arsenicstudies and being a constant source of inspiration during the initial phases of arsenicinvestigations. Thanks are extended to K.K.Singh, A.K.Agrawal, T.B.N.Singh, R.R.Shukla,M. Sonkusare, V.S. Verma, S.Das, Sreehari SMS, S.Upadhyay, S.S.Ganguly, S.K.Singhand K.G.Bhartariya for their support during the arsenic study.

REFERRENCES

• Acharya, S.K. (2005), Arsenic levels in groundwater from Quaternary alluvium in Gangaplain and the Bengal Basin, Indian subcontinent. Insights into influences ofstratigraphy. Gondwana Research, Vol.8 pp. 55-66.
• Allison, G.B.(1982), The relationship between 18O and deuterium in water and sandcolumns undergoing evaporation. J. Hydrol. Vol. 55 pp. 163-176.
• BIS, (2003), Indian standards: drinking water specification (1st revision, Amendment no2). Bureau of Indian Standards, New Delhi
• BGS, (1999), Groundwater studies for arsenic contamination in Bangladesh. PhaseI,Rapid investigation. British Geological Survey and Mott MacDonald Ltd.
• Bhattacharya, P. Chatterjee, D. and Jacks, G. (1997), Occurrence of arseniccontaminatedgroundwater in alluvial aquifers from delta plains, Eastern India: optionsfor safe water supply. Water Resources Development Vol.3 (1) pp. 79-92.
• Breyer, W. (1964), Zur Bestimmung der Wasserdurchla¨ssigkeit von Kiesen und Sandenaus der Kornverteilung. Wasserwirtsch. Wassertech. WWT, Berlin Ost, pp.165–169.
• CGWB, (2006), Groundwater management study in parts of Bhojpur, Buxar and Rohtasdistricts with special reference to arsenic contamination of groundwater, CentralGround Water Board, Mid- Eastern Region, Patna, pp. 46.
• CGWB and BARC (2009), Studies on arsenic pollution of groundwater using isotopicand geochemical methods in arsenic Bhojpur district of Bihar, India, Central GroundWater Board, Mid Eastern Region, Patna pp 49.
• Chakroborty, C. Chattopadhyay, G.S. (2001), Quaternary Geology of South Ganga Plain inBihar. Indian Minerals, Vol. 55(3&4) pp. 133-142.
• Chakraborti, D. Biswas, B. K. Chowdhury, T. R. Basu, G. K. Mandal, B. K. and Chowdhury,U. K. (1999), Arsenic groundwater contamination and sufferings of people in Rajnandgaon,Madhya Pradesh, India. Current Science, Vol.77 pp. 502-504.
• Chakraborty, D. Mukherjee, S.C. Pati, S. Sengupta, M.K. Rahman, M.M. Chowdhury, U.K.Lodh, D. Chanda, C.R. and Chakraborty, A.K. (2003), Arsenic groundwater contamination inMiddle Ganga Plain, Bihar, India: A future Danger? Environmental Health Perspective, Vol.111 pp.1194-1200.
• Chatterjee, D. Roy, R.K. and Basu, B.B. (2005), Riddle of arsenic in groundwater of BengalDelta Plain – role of non-inland source and redox trap. Environ. Earth Science, DOI10.1007/s00257-005-0011-5.
• Driscoll, F.G, (1986) Groundwater and wells. 2nd edition, Johnson Division. St Paul,Minnesota. pp 1089 .
• Gaus, I. Kinniburgh, D.G. Talbot, J.C. and Webster, R. (2003), Geostatistical analysis ofarsenic concentration in groundwater in Bangladesh using disjunctive kriging. EnvironGeol 44:939–948.hronic arsenic poisoning from tube well water. J Indian MedicalAssociation. Vol. 82 pp.34-35
• Guler, C. Thyne, G.D. McCray, J.E. and Turner, A.K. (2002), Evaluation of graphicaland multi-variate statistical methods for classification of water chemistry data.Hydrogeol Journal, Vol.10 pp.455–474.
• Hantush, M.S. and Jacob, C.E. (1955), Non-steady radial flow in an infinite leaky aquifer.Trans. Amererican Geophysical Union, Vol. 36 pp.95-100.
• Kumar, K. Ramanathan, A.L. Rao, M.S. and Kumar, B. (2006), Identification andevaluation of hydrogeochemical processes in the groundwater environment of Delhi,India. Environ Earth Sci. Vol.50 pp.1025–1039.
• Lambe, T.W. (1958), “The structure of compacted clay”. Journal of Soil Mechanics andFoundation Engineering Division, ASCE. 84, 1-35.
• Lee, J.Y. Cheon, J. Y. Lee, K. K. Lee, S. Y. and Lee, M. H. (2001), Statistical evaluationof a geochemical parameter distribution in a groundwater system contaminated withpetroleum hydrocarbons. Jour. Environ Qual, Vol.30 pp.1548–1563.
• Morris, D.A. Johnson, A.I. (1967), Summary of hydrologic and physical properties ofrock and soil material as analyzed by the hydrologic laboratory of the US GeologicalSurvey 1948-1960. USGS Water Supply Paper, pp 1839-D 42.
• Mukherjee, A. Sengupta, M.K. Hossain, M A. Ahmed, S. Das, B. Nayak, B. Lodh, D.Rahman, M. M. and Chakraborty, D. (2006), Arsenic contamination in groundwater: Aglobal perspective with emphasis on the Asian Scenario. J Health Population Nutrition,Vol.24(2) pp.142-163.
• Piper, A. M. (1944), A graphic procedure in the geochemical interpretation of wateranalysis. Am Geophys Union Trans, Vol. 25 pp.914- 923.
• Ravenscroft, P. Burgess, W.G. Ahmed, K.M. Burren M. and Perrin J. (2005), Arsenic ingroundwater of the Bengal Basin, Bangladesh:Distribution, Field relation, andhydrogeological setting, Hydrogeology Jour, Vol.13 pp. 727-751
• Saha, Dipankar. Sreehari, S. Dwivedi, S.N. and Bhartariya, K.G. (2009), Evaluation ofhydrogeochemical 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.
• Singh, I.B. (1996), Geological evolution of Ganga plain – an overview. Jour. Paleont. Soc.India, Vol.1 pp.99-137.
• Singh, I. B. (2004), Late Quaternary history of the Ganga Plain. J. Geol. Soc. India Vol.64 pp 431-454.
• Smedley, P.S (2005) Arsenic occurrence in groundwater in south and east asia-scale,causes and mitigation, in: Towards a More Effective Operational Response-ArsenicContamination of Groundwater in South and East asian Countries, Water andSanitation Programme, World Bank.
• Smedley, P.L. Zhang, M. Zhang, G. and Luo, Z. (2003), Mobilization of arsenic and othertrace metals in fluviolacustrine aquifers of the Huhhot Basin, Inner Mongolia: AppliedGeochemistry, Vol.18 pp.1453–1477.
• Soltan, M.E. (1998), Characterisation, classification and evaluation of somegroundwater samples in upper Egypt. Chemosphere, Vol.37 pp.735–745
• Suk, H. and Lee, K.K. (1999), Characterization of a groundwater hydrochemicalsystemthrough multivariate analysis, clustering into Groundwater zones. Groundwater, Vol.37pp.358–366.
• Tandukar, N. Bhattacharya, P. and Mukherjee, A.B. (2001), Preliminary assessment ofarsenic contamination in groundwater in Nepal In Managing arsenic for ourfuture;proceedings of the International Conference on Arsenic in the Asia-Pacific region,Adelaide, South Australia, November 21-23. Adelaide: Aris Pty. Ltd., pp.103-105.
• WHO. (1993), Guidelines for drinking water quality, 2nd Edn, Vol. 1, World HealthOrganisation, Geneva.
• Walton, W.C. (1962), Selected analytical methods for well and aquifer evaluation. IllionisState Water Survey, Bull. Vol.49 pp-81.

Dipankar Saha, S. N. Dwivedi, Sudarsan Sahu - Central Ground Water Board, MER, Patna