ARTICLE
pubs.acs.org/est
Palaeosol Control on Groundwater Flow and Pollutant Distribution:
The Example of Arsenic
John M. McArthur,† Bibhash Nath,‡,^ Dhiraj M. Banerjee,§ R. Purohit,§ and N. Grassineau||
†
Department of Earth Sciences, University College London, Gower Street, London WC1E 6BT, U.K.
School of Environmental Systems Engineering, The University of Western Australia, M015, 35 Stirling Highway, Crawley,
WA 6009, Australia
§
Department of Geology, University of Delhi, Delhi 110 007, India
Department of Geology, RHUL, Egham, Surrey TW20 0EX, U.K.
)
‡
bS Supporting Information
ABSTRACT: The consumption of groundwater polluted by arsenic (As) has a severe and adverse effect on human health,
particularly where, as happens in parts of SE Asia, groundwater is supplied largely from fluvial/deltaic aquifers. The lateral
distribution of the As-pollution in such aquifers is heterogeneous. The cause of the heterogeneity is obscure. The location and
severity of the As-pollution is therefore difficult to predict, despite the importance of such predictions to the protection of consumer
health, aquifer remediation, and aquifer development. To explain the heterogeneity, we mapped As-pollution in groundwater using
659 wells across 102 km2 of West Bengal, and logged 43 boreholes, to reveal that the distribution of As-pollution is governed by
subsurface sedimentology. Across 47 km2 of contiguous palaeo-interfluve, we found that the shallow aquifer (<70 mbgl) is
unpolluted by As (<10 μg/L) because it is capped by an impermeable palaeosol of red clay (the last glacial maximum palaeosol, or
LGMP, of ref 1) at depths between 16 and 24 mbgl. The LGMP protects the aquifer from vertical recharge that might carry As-rich
water or dissolved organic matter that might drive reduction of sedimentary iron oxides and so release As to groundwater. In 55 km2
of flanking palaeo-channels, the palaeosol is absent, so invasion of the aquifer by As and dissolved organic matter can occur, so
palaeo-channel groundwater is mostly polluted by As (>50 μg/L). The role of palaeosols and, in particular, the LGMP, has been
overlooked as a control on groundwater flow and pollutant movement in deltaic and coastal aquifers worldwide. Models of pollutant
infiltration in such environments must include the appreciation that, where the LGMP (or other palaeosols) are present, recharge
moves downward in palaeo-channel regions that are separated by palaeo-interfluvial regions where vertical recharge to underlying
aquifers cannot occur and where horizontal flow occurs above the LGMP and any aquifer it caps.
’ INTRODUCTION
Pollution of groundwater by naturally occurring arsenic (As)
is found in sedimentary aquifers worldwide.2 The exemplar is
the Bengal Basin, where it is especially widespread and severe in
both Bangladesh3-6 and West Bengal.7-9 The impact on the
health of consumers has been, and remains, considerable and
adverse.10-13
The vertical distribution of As-pollution in shallow groundwater of deltaic aquifers is explicable in terms of sea-level
change1,3,14 and flushing.15-18 It is not clear what causes of the
lateral heterogeneity in the concentrations of As, which ranges
from analytical zero to >1000 μg/L over distances from tens
of meters to kilometres. Understanding such heterogeneity
would help develop strategies for mitigation of As-pollution in,
and exploitation of, deltaic aquifers worldwide that may be Aspolluted. The suggestion that lateral heterogeneity occurs because the pollution is confined to abandoned palaeo-channels
(e.g. ref 8) is supported by the finding1 that, in southern West
Bengal, sedimentology controls the lateral heterogeneity in Aspollution on a scale of a few tens of meters: groundwater from
palaeo-channel aquifers is As-polluted (defined as >50 μg/L As)
and that from palaeo-interfluvial aquifers capped by a palaeosol
r 2011 American Chemical Society
is As-free (defined as <10 μg/L As). This sedimentological control was termed the `palaeosol model’ of As distribution.1 Its key
concept is that palaeo-interfluvial aquifers are capped by an
impermeable palaeosol, formed in the period preceding the last
glacial maximum and so termed the last glacial maximum palaeosol
(LGMP). The LGMP prevents palaeo-interfluvial aquifers from
being recharged by downward movement of overlying groundwaters, which might be rich in As, or might contain dissolved
organic matter that could drive to completion the reduction of
sedimentary iron oxides in the aquifer and so release As to
groundwater. Furthermore, being not much reduced, the sedimentary iron oxides of the palaeo-interfluvial aquifers sorb As
and so keep its concentration in palaeo-interfluvial groundwater
<10, and generally <2 μg/L, while also imparting a brown/
orange color to the sediment.
The palaeosol model proposes that the LGMP, being effectively impermeable,15 is an important control on the direction
Received: September 24, 2010
Accepted: December 22, 2010
Revised:
December 14, 2010
Published: January 26, 2011
1376
dx.doi.org/10.1021/es1032376 | Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
Figure 1. The field area. Base map is 1-in. toposheet of West Bengal circa 1975. Geographic reference system used is WGS84 in degrees and decimal
minutes. Key in upper left. a. Concentrations of As in well water. Solid black line encloses area where As in groundwater is <10 μg/L, and drilling proved
the LGMP to be present. Southern dotted, black, extensions outline the area where erosion has cut out the palaeosol but left underlying Late Pleistocene
brown sands largely undisturbed. b. Drilling sites and lines of traverse for Figure 3.
and rate of groundwater flow and so pollutant transport. Being
global in distribution, the LGMP will therefore control the movement of most soluble pollutants in recharge to deltaic and coastal
aquifers worldwide. The model, if correct, will provide a better
understanding of such pollutant movement with application that
is worldwide. To assess the applicability of the `palaeosol model’
at a scale larger than that studied by McArthur et al. (ref 1)
we tested the model’s predictions across 102 contiguous square
kilometers of southern West Bengal by determining whether
palaeo-interfluvial aquifers in our area are capped by the LGMP
and whether the aquifers beneath any LGMP consist of brown
sand in which groundwater is As-free.
’ METHODS AND DATA
Across 102 km2 of southern West Bengal (Figure 1), we
analyzed for As in 659 well waters, of which 503 were laboratory
analyses. Concentrations of As for 50 other wells were taken from
maps of the Public Health Engineering Department (PHED),
Government of West Bengal (http://www.wbphed.gov.in/
Static_pages/water_quality_survey_reports.html). In the field,
we used a Wagtech Arsenator, supplemented by an Hach As testkit, to analyze groundwater for As. Samples for laboratory
analysis of As were collected in 15 mL polythene tubes and
acidified in the field with 0.15 mL of 50% Analar nitric acid.
Laboratory analysis was done by ICP-AES (samples >50 μg/L
As), hydride-generation ICP-AES (10-50 μg/L As) or ICP-MS
(Agilent 7500 with a He collision-cell <10 μg/L As). On 293
wells, we recorded the color of the chemical precipitate on the well
completions, well, or domestic utensils used for cooking and watercarrying and analyzed the well water for As in order to relate the
color of stain (black for precipitation of manganese oxide, red for
precipitation of iron oxide) to the concentration of As in the
groundwater. We set the expectation that wells tapping groundwater with <10 μg/L As would be Mn-rich and stain black, while
those tapping groundwater containing g10 μg/L would be Fe-rich
and stain red. We also recorded the locations of a further 76 wells
and noted only their stain color and viewed for guidance the color
of many more not recorded. Data on As concentrations in tested
wells are in Table 1. To investigate subsurface sedimentology,
we logged 43 boreholes drilled using the local hand-operated
technique.19 Drill (Table 2) sites were selected primarily after area
surveys of the color of well stains, field analysis of well water for As,
and discussion with drillers cf. von Br€omssen et al. (ref 20).
’ RESULTS AND DISCUSSION
The Palaeosol Model. The distribution of As in groundwater
in the study area is shown in Figure 1a (data in Supporting
Information Table 1), with the locations of drilling sites (Table 2)
in Figure 1b. Within a central area that widens to the east and
1377
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
Figure 2. Sedimentary sequences recovered along the lines of Traverse in Figure 1. Large hand-lettering shows depth in feet of subsamples from
integrated intervals of five feet when unconsolidated, and typical examples of an interval when returned as compact core (which do not represent the full
thickness of the unit cored). Details of typical lithologies shown to the right of main photos. a. Core 7-2009 (Kulberia). Palaeo-interfluvial sequence
from the middle of Traverse 1. A red palaeosol between 60 and 70 feet bgl is overlain by clays and silty clays containing abundant peat and peaty clay (e.g.,
at 30, 35, 45, 50, and 55 feet bgl). The palaeosol (detail to right) at 65 feet bgl shows a carbonized palaeo-rooting structure that confirms its palaeosol
nature. Following ref 1, the palaeosol is termed the last glacial maximum palaeosol (LGMP). A basal clay aquiclude was found at 145 feet bgl. b. Core
3-2009 (Bamangachhi North). Palaeo-channel sequence from the north end of Traverse 1 (Figure 1). Gray sands underlie an upper aquitard of silts and
clays that contain peat at 25 and 35 feet bgl. A basal clay aquiclude was found at 140 feet bgl. c. Core 10-2010 (Kotra). Palaeo-channel sequence from the
southern end of Traverse 2. Gray post-LGM, calcitic sands, overlain by gray silts and silty clays. No organic-rich strata were recovered. A basal clay
aquiclude was found at 145 feet bgl. d. Core 4-2009 (Bara). Shallow palaeo-channel (or truncated palaeo-interfluvial) sequence found centrally south of
the low-As region shown in Figure 1. Post-LGM channelling has removed the palaeosol but not underlying Late Pleistocene, brown, sands on which gray,
post-LGM, sands, and silts lie directly. Drilling stopped at 165 feet, but the driller reported that a basal clay aquiclude was expected at 170 feet bgl.
comprised 47% of the area surveyed, groundwater contained <10
μg/L As (and mostly <2 μg/L), except for three wells northwest
of Chhota Jagulia (88 31.80 E, 22 44.80 N in Figure 1) in which
concentrations of As were 29, 32, and 32 μg/L. In this low-As
region, drilling revealed a palaeo-interfluvial setting in which a
red, oxidized, clay palaeosol overlies a shallow aquifer of brown
sand (Figure 2a). The depth to the top of the palaeosol ranged
from 15 to 24 mbgl (Figure 3). Its thickness was between 1.5 and
6 m and was less in the west (Traverse 1 of Figure 1; mean 2.7 m)
than in the east (Traverse 2 of Figure 1; mean 5.2 m). The
palaeosol is indistinguishable in color, consistency, and structure,
from the LGMP characterized in McArthur et al. (ref 1). At most
sites, the palaeosol contained carbonized rooting holes (e.g.,
Figure 2a) that confirmed it was the LGMP (see ref 1 for further
validation of its palaeosol nature).
The northern margin of the As-free area (where As is <
10 μg/L As) trends northeast/southwest and is transitional to a
northwesterly region where most groundwater contains >50 μg/L
As and much contains >100 μg/L. In this region, drilling revealed
a palaeo-channel sequence of gray, slightly calcitic, sands underlying an upper aquitard of silts and clays that contain peat at
25 and 35 feet bgl (Figure 2b). The LGMP was not found in
this region.
To the southeast and southwest of the As-free area, groundwater is polluted by As (mostly >50 μg/L; Figure 1). In the
southeastern region, at the southern end of Traverse 2, drilling
revealed a typical palaeo-channel sequence of slightly calcitic gray
sands (Figure 2c) underlying silt and silty clay that hosts
common dispersed plant litter. Between these As-polluted
regions, and directly south of the As-free area, concentrations
of As are variable but mostly <50 μg/L. In this region e.g. around
the southern end of Traverse 1, drilling revealed post-LGM,
slightly calcitic, gray, sands overly brown, noncalcitic, sands, with
the contact at around 30 mbgl. We infer that this sequence
represents a truncated palaeo-interfluve (Figure 2d, Figure 3a)
where post-LGM erosion reached a depth sufficient to remove
the LGMP, but not a depth sufficient to cut through much of the
underlying Pleistocene sand, which remains in place and is brown
and iron-oxide bearing. This depth to erosional-base will vary across
the Bengal Basin. In such a condition, groundwater contains mostly
between 10 and 50 μg/L of As because the underlying brown sands
provide sorptive protection against ingress of As, but the hydraulic
protection afforded by the LGMP is lacking.
Our results appear to validate the palaeosol model across the
study area of 102 km2. In the subsurface of the As-free region,
drilling proved the existence of the LGMP capping brown sand
aquifers, as predicted by the model. In flanking regions, where concentrations of As in groundwater mostly exceed 50 μg/L, drilling
proved the absence of the LGMP and that the aquifer consists
mostly of gray palaeo-channel Sands (Figure 3b). In subordinate
1378
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
Figure 3. Sedimentary logs and cross sections along a] Traverse 1 and b] Traverse 2 of Figure 1. The cross sections are schematic but accurately reflect
the regional presence of a basal clay aquitard underlying the shallow aquifer. Palaeo-channel regions are polluted by As. Palaeo-interfluvial areas host
groundwater containing <10 μg/L and mostly <2 μg/L As. At the southern end of Traverse 2, a typical palaeo-channel sequence of brown sand underlies
gray silts and clays containing little organic matter (Figure 2c): here concentrations of As generally exceed 50 μg/L. At the southern end of Traverse 1,
erosion has removed the LGMP but not cut much deeper (Figure 2d), leaving uneroded to sorb As and retard invasion by As-pollution much of the Late
Pleistocene brown sand; in this region, As concentrations are sometimes g10 μg/L but rarely >50 μg/L.
areas, where gray sands directly overly, and are in hydraulic
continuity with, brown sands below (Figure 3a) which sorb As,
pollution by As is minimal. Validation of the model over a larger
area, in other localities, should now be attempted.
Mapping As-Pollution with Well-Color. Our primary guide
in locating areas of palaeo-channel and palaeo-interfluve was the
color of stain on wells, well-completions, and domestic utensils.
Black stain (Figure 4a), from precipitation of manganese oxide,
1379
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
Figure 4. The color of well-stain: a, Black coloration of a well, well-screen, or domestic utensils, derives from precipitation from groundwater of
manganese oxide and indicates that a well is As-free and, in southern West Bengal, also that it is screened in the shallow aquifer. b, Red coloration derives
from precipitation from groundwater of iron oxide and indicates the presence of As g 10 μg/L in shallow wells.
indicated with 90% certainty (see below) that groundwater
tapped by a well would be As-free, and also that the well was
screened in the shallow aquifer (i.e., < 70 mbgl) beneath the
LGMP. Red stain on shallow wells (Figure 4b), from precipitation of iron oxide, indicated with 90% certainty (see below)
that the wells tapped groundwater that was polluted by As. Color
thus revealed rapidly whether a region was polluted or unpolluted by As and allowed us quickly to map palaeo-interfluves
and palaeo-channels. It proved particularly valuable in locating the boundary between them, often to within a few tens
of meters. Color also provided a check in real time on field
analysis for As.
Despite its value, use of color screening requires care. Only
293 (45%) of our wells were clearly colored. Other sampled
wells were either not examined for color (2002 samplings, Table
1) or color was disguised by rust from the pump, algal growth,
regular cleaning to remove color, or colored completions (e.g., of
red brick). Of the 293 clearly colored wells, 109 were judged to be
stained red and expected to contain g10 μg/L As: of these, 11
contained <10 μg/L of As, an error rate of 10%. A further 184
were termed black and expected to be unpolluted (<10 μg/L As),
of which 19 contained g10 μg/L of As and 8 contained >
50 μg/L As, giving error rates of 10% and 4% respectively using
10 and 50 μg/L As to define the term pollution. Some of the
polluted wells were termed black (As-free) because the deepest
red stain resulting from the long-term build-up of thick red stain
on old wells can be confused with black by those challenged by
color vision. In addition, in aquifers with the steep redox gradients typical of West Bengal,1,15 some well-screens intercept
both Mn- and Fe-rich water and so give a mixed color signal.
Finally, groundwater tapped by deep wells (>150 m depth) is
usually Fe-rich and so stains red despite being As-free. It follows
that when using color to screen wells for As-pollution, it is essential to distinguish between deep and shallow wells. Despite
these caveats, the method’s speed and reliability is sufficiently
good for it to be used as a lone screening tool for As-pollution or
its absence in southern West Bengal. Validation of the color of
well-stain as a guide to As-pollution now needs to be tested in
other parts of the Bengal Basin and elsewhere in order to assess
its wider effectiveness.
Use of color when mapping As-pollution will speed prioritization of areas as either at low- or high-risk of As pollution and
allow subsequent well-testing to become more focussed. The
ready availability and low cost of digital cameras permits a permanent record of a well’s state, setting, and color, to be incorporated into a well-record for independent quality control.
Future mapping will require a shift in well-testing away from a
search for As-pollution to a search for its absence. False negatives
in well-testing will replace false positives as a problem in risk
assessment.
Palaeo-Interfluvial Aquifers: Consequences for Water
Supply. Where it is present, the LGMP, which is effectively
impermeable,15 protects the underlying palaeo-interfluvial aquifer from downward migration of mobile pollutants, such as
nitrate, pesticides, NAPLs and DNAPLs, and As, as well as dissolved organic matter (DOM) from the overlying aquitard that
may drive reduction of sedimentary iron oxides and so further
pollute with Fe and As. In our field area, the protection provided
by the impermeable LGMP is strengthened by the presence of
other strata of low permeability that usually overlie it, such as
a pale blue-gray clay (Figure 2a). The LGMP, and other impermeable strata, prevent vertical recharge to palaeo-interfluvial
aquifers and force such potential recharge to flow laterally to
palaeo-channels, from where it flows downward and then laterally
into the palaeo-interfluvial aquifers (Figure 18 of ref 1). This flow
carries with it high concentrations of As. At the margin of the
palaeo-interfluve, the As is stripped from the groundwater by
sorption onto the sedimentary iron oxide in brown sands (our
Figure 5; see also ref 16). The palaeo-interfluves thus act as giant,
in situ, As-removal filters. The retardation factor applicable to Asmigration into the brown sand is uncertain, but modeling and
laboratory studies have suggested values around 140 to 300 for As
concentrations of 900 μg/L.21,22 Values between 30 and 150 were
suggested by McArthur et al. (ref 1) for the palaeo-interfluvial
aquifer in the west of our present study area, although more recent
field monitoring of As migration there implies values between 10
1380
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
Figure 5. Conceptualization of the risk of As-pollution in palaeo-interfluvial aquifers in terms of travel times for As. Note the logarithmic horizontal
scale. Groundwater flow rate is the 30 m per year at palaeo-interfluvial margins given by McArthur et al. (refs 1 and 15) for the western end of the area.
Flow rates decrease inward and are denoted schematically by the size of the horizontal arrows. End-member values for retardation factors, R, are 10, the
minimum likely given by McArthur et al. (ref 16), and 200, toward the maximum of 300 given by Stollenwerk (refs 21 and 22). The impermeable
palaeosol prevents vertical recharge to palaeo-interfluvial aquifers, so recharge moves laterally from flanking palaeo-channels, losing As by sorption as it
does so. Note that the LGMP affords protection to the palaeo-interfluvial aquifer against downward ingress of any mobile pollutants, including nitrate,
pesticide, DNAPL, and NAPL, not just against ingress of natural As-pollution.
and 30:16 much of the range in R may be accounted for by the
operation of scale-factors between laboratory and field and the
likelihood that preferential flow paths exist in aquifers.
Groundwater abstraction occurs across the palaeo-interfluves,
but recharge to the shallow palaeo-interfluvial aquifers can occur
only by lateral inflow at depth from the surrounding palaeochannels. It follows that flow induced by abstraction of water
from the palaeo-interfluvial aquifers is inward at all points around
their margins. This induced flow is superimposed on any regional
flow imposed by natural gradients. The likelihood of As-pollution
reaching a well in a palaeo-interfluvial setting will depend on the
distance of the well from the margin of the palaeo-interfluve
(Figure 5). Wells sited a few tens of meters inward of the margin
may remain unpolluted by As for as little as 10 years (our
Figure 5; see also ref 16). Wells sited more than a hundred meters
downflow of the margin may not be polluted by As for decades.
Our palaeo-interfluve is around 9 km across at its widest point
(Traverse 2, Figure 1b), giving a maximum inflow distance of
possibly half that distance for equant radial inflow. Assuming a
minimum retardation factor of 10 for As transport, the center of
the palaeo-interfluve should be safe from As-pollution for up to
1500 years (Figure 5).
Regular monitoring of sentinel wells along the margins of
palaeo-interfluves should provide early warning of encroaching
As-migration and allow local rates of As-migration to be determined. Such rates may vary between regions in response to
differences in sediment grain size (so sorptive capacity) and the
development of preferential flow-paths. The hydraulic barrier
against downward flow of pollutants (and dissolved organic
matter) provided by the LGMP, coupled with the sorptive barrier
against As invasion provided by the palaeo-interfluvial brown
sands, provide a dual protective barrier against As-pollution that
will allow As-free supplies of water to be developed in palaeointerfluvial locations (e.g., in their centers) where the prognosis
for As-pollution can be made with some confidence (Figure 5).
The LGMP Outside Our Study Area. The potentially shallow
depth (<50 mbgl) of the palaeo-interfluvial aquifers makes
them accessible to local `hand-flap’ drilling methods19 and so
permits well-installations to be made that are affordable to most
of the region’s population. As a consequence, we need to assess
how widely across the Bengal Basin the palaeosol model might
hold, and where it may guide a search for the supply of As-free water.
Some occurrences were documented in Sections 5.7 and 5.8 of
ref 1). Since that publication, the palaeo-interfluvial settings have
been found in parts of Murshidabad, in central West Bengal
(S. Datta, Pers. com. May 2010). Geophysical investigation at
Matlab, in eastern Bangladesh23 appears to reveal palaeo-channels and palaeo-interfluves separated by a “clayey aquitard” that
may be the LGMP (as suggested in ref 1). The palaeosol model
therefore may explain the patchy distribution of As in groundwater in Matlab noted by Jakariya et al. (ref 6) as well as
elsewhere in Bangladesh. In Figure 6 we show a speculative
map of where in Bangladesh such palaeo-interfluvial areas might
be found. Particularly prospective for the LGMP are areas
marginal to the Barind and Madhupur Tracts and those between
the west bank of the palaeo-Brahmaputra River14,24 and the east
bank of a southward flowing palaeo-Ichamati River to the west.
Further afield, the LGMP occurs in the Yangtze Delta25 but has
not yet, to our knowledge, been related to water quality there. As
the LGMP is a generic unit distributed globally, a palaeosol
control may occur globally to significant parts of other deltaic and
coastal aquifers.
Finally, the LGMP is but the last of many ancient land surfaces
that existed across our area as sea-level rose and fell over the past
million years.14 For much of that time, the Barind and Madhupur
Tracts were as upstanding as they are today. The LGMP is thus
but one of many palaeosols that spread across the basin from
these ancient Tracts, in which they coalesce and are condensed,
and from which they resolve, as distance from them increases,
into palaeosols that reflect successive low-stands of sea-level.
Given the prevalence of palaeosols in deltaic settings (ref 26 and
references therein), new finds of the LGMP, and other palaeosols, as a control on the migration of any mobile pollutant,
including As, will not be surprising.
Separation of Deep and Shallow Aquifers. The deep
aquifer has, for decades, been used extensively for irrigation in
1381
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
ARTICLE
unsuitable of many possible places in which to dispose of the spent
sludge from those As-removal cartridges that function by sorbing As
onto iron oxides and are used in the home to reduce As concentrations in drinking water. Disposal would need to be away from the
margins of the palaeo-interfluves (Figure 5) where sorption sites
may be saturated by As or iron oxides reduced by migrating organic
matter or Fe(II) in palaeo-channel groundwater.
Migration of Other Pollutants. Groundwater flow in deltaic
aquifers is controlled, at least in part, by the LGMP in the shallow
subsurface (and possibly other, deeper, palaeosols), except where
post-LGM channelling has removed the LGMP and, to varying
depths, its underlying Late Pleistocene brown sand. This concept
of protection by both LGMP, and underlying, sorptive, brown
sand, provides a model of pollutant control that may be applicable widely across deltaic aquifers to help predict the movement
and distribution of mobile pollutants and contaminants that
originate from the Earth’s surface. The palaeosol model requires
flow-modeling to move away from the idea that pollutants, natural
or anthropogenic, are carried downward in areally extensive recharge and toward a view that downward flow occurs in palaeochannel zones of high permeability where the LGMP is absent
and between which flow is dominantly horizontal in palaeointerfluves above the palaeosol where it will spread pollution
laterally, but not vertically.
Figure 6. Map of the As concentration (μg/L) in shallow groundwater (<150 mbgl) of Bangladesh. Prepared by P. Ravenscroft from
the surveys of DPHE (refs 3 and 4). The areas enclosed in black lines
may be those where the palaeosol may exist at depth in the subsurface,
possibly fragmented by channel erosion. The areas flank major rivers,
major palaeo-rivers,14,24 and the Barind and Madhupur Tracts which
are weathering surfaces of long-standing.
West Bengal and for public supply across the Bengal Basin. Such
long-term abstraction drives concern that pollution from shallow
aquifers may be drawn down to pollute deep aquifers that are
presently free of As.3,27-29 By switching some abstraction for
irrigation from As-free deep aquifers to As-free palaeo-interfluvial
aquifers, a reduction can be achieved in the potential for the
penetration of As to the deep aquifer. The risk is further
mitigated in our area by the presence at most sites drilled of a
basal clay aquiclude that floors the shallow aquifer. Where, for
technical reasons, our drilling stopped short of the anticipated
depth to the clay (<200 feet), its presence was forecast by local
drillers. On this evidence, little connection exists between the
deep and shallow aquifer in our area, but this finding may not
apply elsewhere.
Exploiting Palaeo-Interfluvial Aquifers. Over the past 15
years in our area, small-holders have exploited shallow aquifers in both palaeo-channels and palaeo-interfluves for irrigation in response to the limitations on supply imposed by
aging, government-sponsored, deep irrigation wells. By switching
some of this abstraction from the As-polluted palaeo-channels to
the As-free palaeo-interfluvial aquifers, a reduction could be
achieved in the load of As applied to agricultural soils, in which
the accumulation of As might reduce crop yields.30 In addition,
since the LGM around 20 ka B.P., the palaeo-interfluvial aquifers
have not become sufficiently reducing to mobilize As through
reduction of sedimentary iron oxides to the degree needed to
release As to groundwater. The palaeo-interfluves are also regions of radially inward flow that occurs in a manner that should
be predictable. Their interiors may therefore be the least
’ ASSOCIATED CONTENT
bS
Supporting Information. Table 1 in which As concentrations of well water, well color, and well depth are given with
GPS coordinates to WGS84 datum. Locations of drill sites are in
Table 2. This material is available free of charge via the Internet
at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
Phone: 0044 (0)20 7679 2376; e-mail: j.mcarthur@ucl.ac.uk.
Present Addresses
^
School of Geosciences, The University of Sydney, Sydney,
NSW 2006, Australia.
’ ACKNOWLEDGMENT
This work was funded by the Department of Earth Science,
UCL, a British Council grant to D.M.B. and J.M.M., by a studentship to R.P. from Delhi University, by the University of Western
Australia (B.N.) and by NERC grant NE/G/016879/1. We
thank Peter Ravenscroft for past collaboration and discussions
that were key to the formulation of the palaeosol model developed here, and for an informal review of the final script.
’ REFERENCES
(1) McArthur, J. M.; Ravenscroft, P.; Banerjee, D. M.; Milsom, J.;
Hudson-Edwards, K. A.; Sengupta, S.; Bristow, C.; Sarkar, A.; Tonkin, S.;
Purohit, R. How paleosols influence groundwater flow and arsenic
pollution: A model from the Bengal Basin and its worldwide implication.
Water Resour. Res. 2008, 44, W11411, doi:10.1029/2007WR006552.
(2) Ravenscroft, P.; Brammer, H.; Richards, K. S. Arsenic pollution:
a global synthesis. Wiley-Blackwell, 2009.
(3) DPHE; Groundwater Studies for Arsenic Contamination in
Bangladesh. Final Report, Rapid Investigation Phase. Department of
1382
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383
Environmental Science & Technology
Public Health Engineering, Government of Bangladesh. Mott MacDonald
and British Geological Survey (6 vols), Dhaka, 1999.
(4) DPHE; Arsenic contamination of groundwater in Bangladesh.
Department of Public Health Engineering, British Geological Survey
and Mott MacDonald. BGS Technical Report WC/00/19 (4 vols),
2001.
(5) van Geen, A.; Zheng, Y.; Versteeg, R.; Stute, M.; Horneman, A.;
Dhar, R. K.; Steckler, R.; Gelman, M.; Small, C.; Ahsan, H.; et al. Spatial
variability of arsenic in 6000 tubewells in a 25 km2 area of Bangladesh.
Water Resour. Res. 2003, 39, 1140.
(6) Jakariya, Md.; Vahter, M.; Rahman, M.; Abdul Wahed, M.; Hore,
S. K.; Bhattacharya, P.; Jacks, G.; Persson, L. A. Screening of arsenic in
tubewell water with field test kits: evaluation of the method from public
health perspective. Sci. Total Environ. 2007, 379, 167–175.
(7) PHED; National Drinking Water Mission Project on Arsenic
Pollution in Groundwater in West Bengal. Final Report, Steering
Committee, Arsenic Investigation Project, PHED, Kolkata, Govt. of
West Bengal, India, 1991, 57pp.
(8) Nath, B.; Berner, Z.; Basu Mallik, S.; Chatterjee, D.; Charlet, L.;
Stueben, D. Characterization of aquifers conducting groundwaters with
low and high arsenic concentrations: a comparative case study from
West Bengal, India. Mineral. Mag. 2005, 69, 841–854.
(9) Nickson, R.; Sengupta, S.; Mitra, P.; Dave, S. N.; Banerjee, A. K.;
Bhattacharya, A.; Basu, S.; Kakoti, N.; Moorthy, N. S.; Wasuja, M.; et al.
Current knowledge on the distribution of arsenic in groundwater in five
states of India. J. Environ. Sci. Heal. A 2007, 42, 1707–1718.
(10) Dhar, R. K.; Biswas, B. K.; Samanta, G.; Mandal, B. K.;
Chakraborti, D.; Roy, S.; Jafar, A.; Islam, A.; Ara, G.; Kabir, S.; et al.
Groundwater arsenic calamity in Bangladesh. Curr. Sci. 1997, 73, 48–59.
(11) Smith, A. H.; Lingas, E. O.; Rahman, M. Contamination of
drinking-water by arsenic in Bangladesh: a public health emergency. B.
World Health Organ. 2000, 78, 1093–1103.
(12) Mukherjee, A.; Sengupta, M. K.; Hossain, M. A.; Ahamed, S.;
Das, B.; Nayak, B.; Lodh, D.; Rahman, M. M.; Chakraborti, D. Arsenic
contamination in groundwater: a global perspective with emphasis on
the Asian scenario. Jour. Health Popul. Nutr. 2006, 24, 142–163.
(13) Argos, M.; Kalra, T.; Rathouz, P. J.; Chen, Y.; Pierce, B.; Parvez,
F.; Islam, T.; Ahmed, A.; Rakibuz-Zaman, M.; Hasan, R.; et al. Arsenic
exposure from drinking water, and all-cause and chronic-disease mortalities in Bangladesh (HEALS): a prospective cohort study. Lancet
2010, 376, 252–258.
(14) Umitsu, M. Late Quaternary sedimentary environments and
landforms in the Ganges Delta. Sediment. Geol. 1993, 83, 177–186.
(15) McArthur, J. M.; Banerjee, D. M.; Hudson-Edwards, K. A.;
Mishra, R.; Purohit, R.; Ravenscroft, P.; Cronin, A.; Howarth, R. J.;
Chatterjee, A.; Talukder, T.; et al. Natural organic matter in sedimentary
basins and its relation arsenic in anoxic ground water: the example of
West Bengal and its worldwide implications. Appl. Geochem. 2004, 19,
1255–1293.
(16) McArthur, J. M.; Banerjee, D. M.; Sengupta, S.; Ravenscroft, P.;
Klump, S.; Sarkar, A.; Disch, B.; Kipfer, R. Migration of As, and 3H/3He
ages, in groundwater from West Bengal: implications for monitoring.
Water Res. 2010, 44, 4171–4185.
(17) Klump, S.; Kipfer, R.; Cirpka, O. A.; Harvey, C. F.; Brennwald,
M. S.; Ashfaque, K. N.; Badruzzaman, A. B. M.; Hug, S. J.; Imboden,
D. M. Groundwater dynamics and arsenic mobilization in Bangladesh
assessed using noble gases and tritium. Environ. Sci. Technol. 2006, 40,
243–250.
(18) van Geen, A.; Zheng, Y.; Goodbred, S., Jnr; Horneman, A.; Aziz,
A.; Cheng, Z.; Stute, M.; Mailloux, B.; Weinman, B.; Hoque, M. A.; et al.
Flushing History as a Hydrogeological Control on the Regional Distribution of Arsenic in Shallow Groundwater of the Bengal Basin.
Environ. Sci. Technol. 2008, 42, 2283–2288.
(19) Ali, M. Review of drilling and tubewell technology for groundwater
irrigation. In Groundwater Resources and Development in Bangladesh Background to the Arsenic Crisis, Agricultural Potential and the Environment; Rahman, A. A., Ravenscropt, P., Eds.; University Press Ltd.:
Dhaka, 2003.
ARTICLE
(20) von Br€omssen, M.; Jakariya, M.; Bhattacharya, P.; Ahmed,
K. M.; Hasan, M. A.; Sracek, O.; Jonsson, L.; Lundell, L.; Jacks,
G. Targeting low-arsenic aquifers in Matlab Upazila, Southeastern
Bangladesh. Sci. Total Environ. 2007, 379, 121–132.
(21) Stollenwerk, K. G.; Breit, G. N.; Welch, A. H.; Yount, J. C.;
Whitney, J. W.; Foster, A. L.; Uddin, M. N.; Majumder, R. K.; Ahmed, N.
Arsenic attenuation by oxidized aquifer sediments in Bangladesh. Sci.
Total Environ. 2007, 379, 133–150.
(22) Stollenwerk, K. G. Geochemical implications for sustainable
utilization of low-arsenic aquifers in the Bengal Basin. Proceedings,
Goldschmidt 2010, June 13-18, 2010. Geochim. Cosmochim. Ac. 2010,
74 (Supplement 1), A995.
(23) Hossain, M.; Islam, M. M.; Rashid, S. M. A.; Rahman, M. M.;
Rahman, M.; Sultana, S.; Ahmed, K. M.; Hasan, A.; Bhattacharya, P.;
Jacks, G.; et al. Delimiting the shallow aquifer characteristics using
vertical electrical soundings (VES) and hydrochemical variability in a
region with high arsenic groundwater in southeastern Bangladesh. In
Third International Congress: Arsenic in Geosphere and Human Diseases;
Jean, J. -S., Bhattacharya, P., Bundschuh, J., Eds.; CRC Press, Taylor and
Francis, 2010, 105-107.
(24) Ravenscroft, P.; McArthur, J. M. Mechanism of pollution of
groundwater by boron: the examples of Bangladesh and Michigan, USA.
Appl. Geochem. 2004, 19, 1413–1430.
(25) Li, C.; Wang, P.; Sun, H.; Zhang, J.; Fan, D.; Deng, B. Late
Quaternary incised-valley fill of the Yangtze delta (China): its stratigraphic framework and evolution. Sediment. Geol. 2002, 152, 133–158.
(26) Kraus, M. J. Paleosols in clastic sedimentary rocks: their
geologic applications. Earth-Sci. Rev. 1999, 47, 41–70.
(27) Mukherjee, A.; Fryar, A. E.; Howell, P. D. Regional hydrostratigraphy and groundwater flow modeling in the arsenic-affected areas
of the western Bengal basin, West Bengal, India. Hydrogeol. J. 2007, 15,
1397–1418.
(28) Michael, H. A.; Voss, C. I. Controls on groundwater flow in the
Bengal Basin of India and Bangladesh: regional modeling analysis.
Hydrogeol. J. 2009, 17, 1561–1577.
(29) Burgess, W. G.; Hoque, M. A.; Michael, H. A.; Voss, C. I.; Breit,
G. N.; Ahmed, K. M. Vulnerability of deep groundwater in the Bengal
aquifer system to contamination by arsenic. Nature Geosci. 2010, 3,
83–87.
(30) Brammer, H.; Ravenscroft, P. Arsenic in groundwater: a threat
to sustainable agriculture in South and South-east Asia. Environ. Int.
2009, 35, 647–654.
1383
dx.doi.org/10.1021/es1032376 |Environ. Sci. Technol. 2011, 45, 1376–1383