Three-dimensional mathematical model to simulate groundwater flow in the lower Palar River basin, southern India M. Senthilkumar · L. Elango
Abstract A three-dimensional mathematical model to simulate regional groundwater flow was used in the lower Palar River basin, in southern India. The study area is characterised by heavy abstraction of groundwater for agricultural, industrial and drinking water supplies. There are three major pumping stations on the riverbed apart from a number of wells distributed over the area. The model simulates groundwater flow over an area of about 392 km2 with 70 rows, 40 columns, and two layers. The model simulated a transient-state condition for the period 1991–2001. The model was calibrated for steady- and transient-state conditions. There was a reasonable match between the computed and observed heads. The transient model was run until the year 2010 to forecast groundwater flow under various scenarios of overpumping and less recharge. Based on the modelling results, it is shown that the aquifer system is stable at the present rate of pumping, excepting for a few locations along the coast where the groundwater head drops from 0.4 to 1.81 m below sea level during the dry seasons. Further, there was a decline in the groundwater head by 0.9 to 2.4 m below sea level in the eastern part of the area when the aquifer system was subjected to an additional groundwater withdrawal of 2 million gallons per day (MGD) at a major pumping station. Rsum Les modles mathmatiques en trois dimensions de l’coulement souterrain rgional sont trs utiles pour la gestion des ressources en eau souterraine, car ils permettent une valuation des composantes des processus hydrologiques et fournissent une description physique de l’coulement de l’eau dans un aquifre. Une telle modlisation a t entreprise sur une partie du bassin infrieur de la rivire Palar, dans le sud de l’Inde. La zone d’tude est caractrise par des prlvements importants Received: 9 September 2002 / Accepted: 9 September 2003 Published online: 16 January 2004
Springer-Verlag 2004 M. Senthilkumar · L. Elango ()) Department of Geology, Anna University, 600025 Chennai, India e-mail:
[email protected] Hydrogeology Journal (2004) 12:197–208
d’eau souterraine pour l’agriculture, l’industrie et l’eau potable. Il existe trois grandes stations de pompage sur la rivire en plus d’un certain nombre de puits rpartis dans cette rgion. Le modle simule l’coulement souterrain dans une rgion d’environ 392 km2 avec 70 rangs, 40 colonnes et deux couches. Le modle a fonctionn en rgime transitoire en utilisant une approximation aux diffrences finies d’une quation diffrentielle partielle en trois dimensions de l’coulement souterrain dans cet aquifre pour la priode 1991–2001. Le modle a t calibr pour des conditions de rgime permanent et transitoire. Les charges hydrauliques calcules taient en bon accord avec celles observes. Sur la base des rsultats du modle, il est apparu que le systme aquifre est stable pour ce taux de pompage, except en quelques sites le long de la cte o l’eau marine a pntr 50–100 m dans les terres. Le modle transitoire a tourn jusqu’en 2010 afin de prvoir l’coulement souterrain dynamique pour diffrents scnarios de pompage excessif et de recharge rduite. Il se produit un abaissement de la pizomtrie de la nappe de 0.6 0.8 m dans la partie orientale, alors que l’aquifre est soumis un prlvement supplmentaire de 8,000 m3/jour l’une des stations principales de pompage. MÞme avec le niveau actuel de pompage, la pizomtrie de la nappe descendrait sous le niveau de la mer au cours des saisons sches. Le modle prdit le fonctionnement du systme aquifre sous diffrentes conditions de stress hydrologique. Resumen Los modelos tridimensionales de flujo de aguas subterrneas son ffltiles para gestionar los recursos hdricos subterrneos, ya que proporcionan una aproximaci n a los diversos procesos hidrol gicos y una descripci n cuantitativa del flujo de agua en el acufero. Se ha desarrollado un estudio de modelaci n de este tipo en una parte de la cuenca baja del ro Palar, en el Sur de la India. Esta zona se caracteriza por las intensas extracciones de aguas subterrneas para usos agrcolas, industriales y domsticos. Hay tres estaciones de bombeo principales en el ro, adems de numerosos pozos distribuidos por la zona. El modelo simula el flujo de las aguas subterrneas en una superficie de 392 km2 por medio de 70 filas, 40 columnas y 2 capas. El modelo ha sido empleado en condiciones transitorias, por medio de la aproximaci n en diferencias finitas de las ecuaciones diferenciales parciales en tres dimensiones del flujo en el acufero durante el DOI 10.1007/s10040-003-0294-0
198
perodo 1991–2001. Se ha calibrado el modelo en condiciones permanentes y transitorias. El ajuste entre los niveles calculados y medidos es razonable. A partir de los resultados de la modelaci n, se ha obtenido que el sistema acufero es estable con la tasa de bombeo utilizada, exceptuando unos pocos emplazamientos a lo largo de la costa, donde se ha dado lugar a fen menos de intrusi n marina en una distancia de 50–100 m. El modelo transitorio ha sido ejecutado hasta el a o 2010 para predecir el flujo dinmico bajo diversos escenarios de sobreexplotaci n y de reducci n de la recarga. Se produce una disminuci n en los niveles piezomtricos de 0.6 a 0.8 m en la zona oriental, donde el sistema acufero est sometido a una extracci n adicional de 2 millones de galones por da en la estaci n principal de bombeo. Incluso con las extracciones actuales, los niveles piezomtricos se sitfflan bajo el nivel del mar durante las pocas secas. El modelo predice el comportamiento de este sistema acufero bajo varias condiciones de presi n hidrol gica. Keywords Numerical modelling · Coastal aquifer · Groundwater flow · Groundwater management · Palar River basin
Introduction Increasing demand for groundwater due to ever increasing populations has initiated the need for and effective management of available groundwater resources. Groundwater modelling is a powerful management tool which can serve multiple purposes such as providing a framework for organising hydrologic data, quantifying the properties and behaviour of the systems, and allowing quantitative prediction of the responses of those systems to externally applied stresses. A three-dimensional groundwater model is an empirically effective tool. A number of groundwater modelling studies have been carried out around the world for effective groundwater management (Corbet and Bethke 1992; Strom and Mallory 1995; Gomboso et al. 1996; Strom 1998; Gnanasundar and Elango 2000; Senthilkumar and Elango 2001; Selroos et al. 2002). Such a study was attempted for the lower Palar River basin located in southern India. As the Palar River flows only for a few days in a year, groundwater has been extensively used to meet the increasing demand for domestic, irrigational and industrial requirements. Industrial abstraction includes pumping for the Madras atomic power station (MAPS). Due to the dependence on groundwater for all these purposes, a groundwater model will help to effectively manage the aquifer system. Hence, the present study was carried out with the objective of developing a regional groundwater flow model which will aid in the management of groundwater in the lower Palar River basin.
Hydrogeology Journal (2004) 12:197–208
Study Area The part of the lower Palar River basin, Tamil Nadu, India, considered for this study is located 75 km south of Chennai (formerly Madras) and covers an area of 392 km2 (Fig. 1). The eastern side of this area is bounded by the Bay of Bengal. The area enjoys a subtropical monsoon climate, with January and February as the dry periods and March–May as the summer period, followed by the monsoon period. The maximum temperature in this area is about 42 C during the months of May and June. The minimum temperature is about 21 C, recorded during the months of December and January. The southwest monsoon (June–September), the northeast monsoon (October– December) and the transition period contribute 40, 51 and 9% respectively of the total average annual rainfall (1,167 mm/year) measured at the two rainfall stations (Fig. 2). The area is approximately bisected by the Palar River. This is a seasonal river flowing during the months of November, December and January. Numerous ponds are present in the depressed parts of the undulating topography.
Hydrogeology The study area exhibits varied physiographic features with elevation ranging from 40 m in the west to sea level in the east. Geologically, the area has two distinct formations: crystalline rocks of Archaean age and recent alluvium. The alluvial deposits occur along the present and palaeo-Palar River courses. Crystalline rocks comprising charnockites and gneiss form the basement and some exposures are found in the southern part of the area. Alluvium occurs as the upper layer and is characterised by sand, gravel and sandy clay. Its thickness ranges from 1 m at the northern and southern boundaries to 30 m along the river. The alluvium and weathered crystalline charnockites function as an unconfined aquifer system. Groundwater occurs under unconfined conditions in both the alluvial and underlying weathered rocks. Depth to groundwater from local monitoring wells (Fig. 2) varies in the range 2–5.5 m below ground level. The hydraulic conductivity of the alluvium is 20–69 m/day. The specific yield value varies in the range 0.037–0.32 (PWD 2000). The lower layer is comprised of weathered crystalline rocks with thicknesses of 0–7 m. The hydraulic conductivity of this layer is 0.5–12 m/day, and specific yield values are 0.002–0.01 (PWD 2000).
Model Formulation The conceptual model of the hydrogeologic system was derived from a detailed study of the geology, borehole lithology and water-level fluctuations in wells. Groundwater of the study area was found to occur in both alluvial formations and in the underlying weathered rocks. The measured groundwater head in monitoring wells, screened in the alluvial and weathered formation, is about the same. Based upon this information, the alluvial and the DOI 10.1007/s10040-003-0294-0
199
Fig. 1 The lower Palar River basin, southern India
Fig. 2 Monitoring well and data site location map of the study area Hydrogeology Journal (2004) 12:197–208
DOI 10.1007/s10040-003-0294-0
200
Fig. 3 Discretisation of the study area
weathered fractured rocks can be considered as a single unconfined aquifer consisting of two sublayers, the upper alluvial layer (sand, sandy clay, clay) and the lower weathered rock layer and the aquifer was conceptualised as an unconfined two-layered system.
cells is 500 m along both the east–west and north–south directions of the study area. Vertical cross sections of the system along A–A0 and B–B0 are shown in Fig. 4.
Input Parameters
Boundary Conditions
Aquifer Characteristics
The northern, western and southern boundaries of the study area are the watershed boundary formed by massive charnockite. The flow from these boundaries into the system is negligible and hence they are considered as noflow boundaries (Fig. 3). About 4 km in the north-western part of the study area is considered as a variable head boundary (Fig. 3), as this region has alluvial deposits of about 10-m thickness. The eastern part of the study area is bounded by the Bay of Bengal, which is taken as a constant head boundary. The aquifer top and bottom were derived mainly from the lithology of 12 boreholes (PWD 2000) and by intensive field surveys. The unconfined aquifer is divided into two layers, the upper alluvial layer and the lower weathered rock layer. The thickness of the upper sublayer varies in the range 0–31 m, with maximum thickness along the sides of the Palar River. The thickness of the lower sublayer is 0–7 m.
The aquifer properties, such as horizontal hydraulic conductivity (kh) and vertical hydraulic conductivity (kv), used in the model were derived from eight pumping test (Fig. 2) results (PWD 2000), and are given in Table 1.
Grid Design The model grid covering 392 km2 of the study area was discretised into 4,800 cells with 70 rows and 40 columns, and vertically by two layers (Fig. 3). The length of model Hydrogeology Journal (2004) 12:197–208
Groundwater Abstraction The groundwater of the study area is abstracted for irrigation, industrial and domestic purposes. Agriculture is the main activity of the study area. The land use of the study area includes about 210 km2 of irrigational activities, of which 147 km2 is dependent on groundwater. The groundwater abstraction for irrigational activities was calculated based on the total cropland, yield of the wells and hours of pumping. Industrial pumping includes 3.5 MGD (PWD 2000) for the Madras atomic power station (MAPS) from its pumping station on the Palar River (Fig. 2). Another pumping station, which supplies water to other industries, is located in the Ayapakkam village which pumps about 0.75 MGD (PWD 2000). Another pumping station located at Vallipuram supplies about 0.5 MGD (PWD 2000) of drinking water for the Chennai city outskirts. Apart from these, domestic DOI 10.1007/s10040-003-0294-0
201
Fig. 4 Cross section along A–A0 (left), and cross section along B–B0 (right)
Table 1 Pumping test results (after PWD 2000) Well
Name of village
Lithology
K (m/day)
S
P1 P2 P3 P4 P5 P6 P7 P8
Paiganallur Pillappur Ayapakkam Issur Voyalur Manapakkam Madurantakam Pallipattu
GL–21 m: sand, gravels with pebbles GL–11.6 m: sand and clay; 11.6–16.7 m: sand GL–4 m: fine to coarse sand; 4–12 m: silt; 12–19.8 m: clayey silt GL–1.5 m: granular zone; 1.5–9 m: weathered charnockite GL–7 m: coarse sand; 7–11 m: clay sand; 11–23 m: sand, gravel with pebbles GL–1 m: top soil; 1–5 m: Kankar; 5–17 m: weathered charnockite GL–1.5 m: top soil; 1.5–5 m: sandy clay; 5–7.5 m: weathered charnockite GL–4.5 m: fine to coarse sand; 4.5–10.8 m: sand, gravel
54 40 69 8 61 10 12 48
0.32 0.228 0.348 0.012 0.322 0.01 0.01 0.228
pumping for household needs was calculated to be 0.3 MGD based on population which was assigned to cells covering the regions of settlement.
Groundwater Recharge The recharge to the aquifer varies considerably due to differences in land-use pattern, soil type, topography and relief. Recharge is from rainfall, irrigation return flow, and inflow from the river and ponds. Rainfall is the principal source of groundwater recharge. The study area was divided into five zones—zones A, B, C, D and E based on rock and soil types. The recharge values incorporated in the model for each zone are shown in Fig. 5. Maximum recharge occurs along the banks of the Palar River, i.e. zone A. A comparison between the monthly rainfall value and consequent variation of groundwater level for a span of 30 years revealed that groundwater is replenished whenever the monthly rainfall exceeds 60 mm. The rate of leakage between the river and aquifer was estimated using the difference between the river head and groundwater head. The contribution of the Killiyar River, a passage canal from Madurantakam Lake to the Palar River, to recharge was also considered. Numerous ponds are present in the study area (Fig. 1). The recharge from the ponds was estimated from the Hydrogeology Journal (2004) 12:197–208
difference between the pond water head and groundwater head. Water is available only during the months of rainfall in almost all ponds of the area. Hence, recharge from these ponds was calculated by considering a water level of 50 cm above ground level during the months when the rainfall exceeds 300 mm, as no pond water-level data are available. In the case of Madurantakam Lake, which is the only perennial lake of the study area, the recharge rates were calculated using the monthly lake water-level data.
Model Description Anisotropic and heterogeneous three-dimensional flow of groundwater, assumed to have constant density, described by the partial differential equation given by Rushton and Redshaw (1979), was used to model the groundwater flow in the study. The finite-difference computer code MODFLOW (McDonald and Harbaugh 1998), which numerically approximates this equation, was used to simulate the groundwater flow in the study area. The pre- and postprocessor, developed by the United States Department of Defense Groundwater Modelling System (GMS), was used to give input data and process the model output.
DOI 10.1007/s10040-003-0294-0
202
Fig. 5 Rainfall recharge zones and amount of recharge in the study area
Table 2 Initial and calibrated hydraulic parameters
S. no.
1 2 3
Geology of the area
Sand Sandy clay Weathered charnockite
Model Calibration The calibration strategy was to initially vary the bestknown parameters as little as possible, and vary the poorly known or unknown values the most to achieve the best overall agreement between simulated and observed datasets. Steady-state model calibration was carried out to minimize the difference between the computed and field water-level conditions, with the water-level data of January 1991 in 17 wells distributed over the study area. Out of all the input parameters, the hydraulic conductivity value is the only poorly known one, as only eight pumping tests have been carried out in this area. The lithological variations in the area and borehole lithology of existing large-diameter wells were studied. Based on these data, it was decided to vary hydraulic conductivity values up to 10% of the pumping test results for both the upper and lower layers in order to get a good match of the computed and observed heads (Fig. 6). The figure indicates that there is a very good match between the calculated and observed heads in most of the wells of the
Hydrogeology Journal (2004) 12:197–208
Hydraulic conductivity (m/day)
Specific yield
Initial
Calibrated
Initial
Calibrated
70 37 12
76 32 7
0.29 0.18 0.02
0.34 0.24 0.03
study area. The root mean square error and the mean error were minimized through numerous trial runs. Transient-state simulation was carried out for a period of 11 years from January 1991 to December 2001, with monthly stress periods and 24-h time steps. The trial and error process by which calibration of the transient model was achieved included several trials until a good match was achieved between computed and observed heads over space and time. The hydraulic conductivity values incorporated in the transient model were modified slightly from those calibrated by the steady-state model. Table 2 gives the initial and calibrated hydraulic conductivity arrived at after calibration. Based on the close agreement between measured and computed heads from January 1991 to December 2001 at 17 observation wells (Fig. 2) distributed throughout the aquifer, the transient models were considered to be calibrated satisfactorily (Fig. 6). The sensitivity of the model to input parameters was tested by varying only the parameter of interest over a range of values, and monitoring the response of the model by determining the root mean square error of the simulated heads compared to the measured heads. These DOI 10.1007/s10040-003-0294-0
203 Fig. 6 Comparison of computed and observed groundwater head (m) under steady-state calibration in January 1991 (top), and comparison of computed and observed groundwater head (m) under transientstate calibration (bottom)
analyses showed that the model is most sensitive to recharge.
Simulation Results The model was simulated in transient condition for a time interval of 11 years in the period 1991–2001. There was fairly good agreement between the computed and observed heads (Fig. 7). A study of the simulated potentiometric surface of the aquifer indicates that the highest heads are found on the western side of the study area, which is a general reflection of the topography. The regional groundwater flow direction is towards the Hydrogeology Journal (2004) 12:197–208
southeast. The simulated groundwater flow vectors also clearly indicate the inward movement of seawater along this coastal region near the villages of Chinnakuppam, Periyakuppam and Oilyakuppam. The simulated and the observed regional heads for the stress period 3800 (September 2001) are shown in Fig. 7. The computed and observed heads in well nos. 4 and 6 of the study area are shown in Fig. 8. The computed head values mimic the observed head values. There is a very gradual decline in groundwater head over ten years.
DOI 10.1007/s10040-003-0294-0
204
Fig. 7 Simulated groundwater head for September 2001 (top), and observed groundwater head for September 2001 (bottom)
Model Forecast The aquifer response for different input and output fluxes was studied in order to sustainably manage the lower Palar River basin aquifer system. The model was run for a further period of nine years from 2001 to 2010. Before Hydrogeology Journal (2004) 12:197–208
commencement of this simulation, the data of average rainfall, abstraction, pond water, river flow and recharge up to 2010 were put into the model. Analysis of the last ten years of river flow data indicated that the Palar River flowed only once in three years. Hence, the Palar River
DOI 10.1007/s10040-003-0294-0
205 Fig. 8 Simulated and observed head at well nos. 4 and 6
The model was run to predict the regional groundwater head in this area until the year 2010. For these runs the monthly average rainfall calculated from 60 years of rainfall data was used The present level of groundwater abstraction was considered for this simulation. The simulated regional groundwater head for September 2010 is shown in Fig. 9. There was not much increase or decrease in water level (Fig. 10). In the years when the flow in the Palar River was considered, there is an increase in groundwater level by about 0.5 m in the wells located near the rivers.
1,167 mm/year. The average of these low-rainfall years (drought period) was found to be 882.43 mm/year. In order to study the effect of drought years in this area, the model was run by assuming deficit rainfall once in four years until 2010. The monthly average of deficitrainfall years was calculated and used for this purpose. Groundwater levels declined by about 0.6–1.2 m during the assumed drought years (Fig. 10). However, the groundwater level recovered to the level observed during the normal rainfall within the next year. Even during normal-rainfall years, water levels ranged from 0.4 to
1.8 m below sea level in wells located up to 800 m from the coast. The groundwater head lowers by an additional
0.6 m below sea level during drought years in the coastal region. During these drought periods there would be an increase in pumping, which was not considered in this simulation. In the case of excessive pumping during such periods, seawater intrusion will occur in the coastal areas.
Drought Year Once in Four Years
Increase in Pumping
Analysis of the past 60 years of rainfall data indicates that in 33 years the rainfall was less than the average of
Groundwater is the major source of water for industries located in the project area and there has been an increase
flow was considered once in three years, that is, during 2003, 2006 and 2009. Three prediction runs were planned to evolve optimal management schemes.
Normal Rainfall Condition
Hydrogeology Journal (2004) 12:197–208
DOI 10.1007/s10040-003-0294-0
206 Fig. 9 Simulated groundwater head in September 2010
Fig. 10 Observed and predicted groundwater head until December 2010 under normal conditions (top), and predicted groundwater head with decrease in rainfall once in four years and under present conditions (bottom)
in pumping over the years. Hence, it is essential to know the behaviour of the system under increased hydrological stress. There are two major, anticipated changes in the pumping pattern in the future, the first being the possible increase in pumping at the MAPS pumping station to meet the increasing demand for the expansion of its activities, and the second being the increase in pumping for various uses over the entire area. These two situations are discussed below.
Hydrogeology Journal (2004) 12:197–208
Increase in pumping by 15% at the MAPS pumping station It was anticipated that pumping will be increased by 2 MGD for the proposed prototype Fast Breeder Reactor near MAPS. Hence, the model was run with an increase of 2 MGD (15% increase) in pumping at the station. For these runs, the monthly average rainfall calculated from the 60 years of rainfall data was used. The predicted regional groundwater head with increase in pumping is shown in Fig. 11. In well no. 6 located in Pandur Village (located at the western side of the pumping station), the groundwater head is lowered by 0.4–0.6 m due to an DOI 10.1007/s10040-003-0294-0
207 Fig. 11 Predicted groundwater head for September 2010 with increase in pumping by 2 MGD at the MAPS site and under normal pumping conditions
Fig. 12 Groundwater head with present pumping conditions and with an increase in pumping by 2 MGD (15% at MAPS) at well no. 6 (western side of MAPS site) (top), groundwater head with present conditions and with an increase in pumping by 2 MGD (15% at MAPS) at well no. 4 (eastern side of MAPS site) (middle), and groundwater head with present pumping conditions and with an increase in pumping by 2 MGD (15% overall increase) at well no. 6 (bottom)
Hydrogeology Journal (2004) 12:197–208
DOI 10.1007/s10040-003-0294-0
208
increase in pumping (Fig. 12). In well no. 4 located in Lathur Village (located at the eastern side of the pumping station), the groundwater head is lowered by 0.6–0.8 m due to an increase in pumping (Fig. 12). The comparison between the wells located in the western and eastern parts of the MAPS site indicates that the groundwater level decreases more at the eastern side. Even under the normal rate of pumping, the groundwater head is lowered below sea level during the dry seasons (as discussed earlier) whereas, due to the increase in pumping at the MAPS pumping station, the groundwater head would decline much lower than sea level. The flow vectors also indicate that about 2–2.5 km inland from the coast would be affected by saline intrusion, resulting in contamination of groundwater in the area. Thus, if pumping is increased by 2 MGD at the MAPS pumping station, the villages in the eastern part of the basin will be affected. Increase of pumping by 15% in the entire study area The simulated groundwater pumping rate for the entire study area was increased by an additional 2 MGD (i.e. 15% increase). For these runs, the monthly average rainfall calculated from the 60 years of rainfall data was used. The forecasted model indicates that this aquifer system is stable and has no adverse effect due to the increase in pumping. The groundwater head value was found to be lowered by only 0.1–0.3 m (Fig. 12). This clearly indicates that the aquifer is stable but an increase of 2 MGD pumping uniformly distributed over the entire area does affect the aquifer system.
Conclusion A two-layered finite-difference flow model was used to simulate the groundwater head in the lower Palar River basin for a period of 11 years (1991–2001) for a better understanding of the aquifer system. The simulated results indicate that this aquifer system is stable under the present pumping rate, except for a few locations along the coast where seawater would intrude up to 50–100 m inland. The model predicted that an increase of 2 MGD pumping (15% increase) at the MAPS pumping site would lower the groundwater head by 0.6–0.8 m on the eastern side and 0.4–0.6 m on the western side of this
Hydrogeology Journal (2004) 12:197–208
pumping station. Even with the present level of pumping, the groundwater head is lowered below sea level along the coast during the dry seasons and, if the pumping is increased at the MAPS pumping station, the groundwater head would decline by 0.9–2.4 m below sea level in the eastern part of the area. Acknowledgement Groundwater-level data provided by the Public Works Department, Tamil Nadu is acknowledged. The authors thank Dr. N. Rajeshwara Rao, Lecturer, Dept. of Applied Geology, Madras University for his help in correcting this manuscript. The authors wish to thank two anonymous reviewers for their comments.
References Corbet TF, Bethke CM (1992). Disequilibrium fluid pressures and groundwater flow in western Canada sedimentary basin. J Geophys Res 97(B5):7203–7217 Gnanasundar D, Elango L (2000) Groundwater flow modeling of a coastal aquifer near Chennai City, India. J Indian Water Resources Soc 20(4):162–171 Gomboso J, Ghassemi F, Jakeman AJ (1996) Modeling groundwater flow in the Northern Stirling land conservation district western Australia. Ecol Model 80:169–175 McDonald MG, Harbaugh AW (1998) User’s documentation for MODFLOW-98, an update to the U.S. Geological Survey modular finite-difference groundwater flow model. US Geol Surv Open-File Rep 96-485, 56 pp PWD (2000) Groundwater perspectives. A profile of Kancheepuram district, Tamil Nadu. Public Works Department, Tamil Nadu, June, 220 pp Rushton KR, Redshaw SC (1979) Seepage and groundwater flow. Wiley, New York, 330 pp Selroos JO, Walker DD, Strom A, Gylling B, Follin S (2002) Comparison of alternative modelling approaches for groundwater flow in fractured rock. J Hydrol 257(1/4):174–188 Senthilkumar M, Elango L (2001) Numerical simulation of groundwater flow regime in a part of the lower Palar River basin, southern India. In: Elango L, Jayakumar R (eds) Modelling in hydrogeology. UNESCO-IHP Allied Publ, 270 pp Strom EW (1998) Hydrogeology and simulation of groundwater flow in the Cretaceous-Paleozoic aquifer system northern Mississippi. US Geol Surv Water-Resources Inv Rep 984171, 81 pp Strom EW, Mallory MJ (1995) Hydrogeology and simulation of groundwater flow in the Eutaw-Mcshan aquifer and in the Tuscaloosa aquifer system in northeastern Mississippi. US Geol Surv Water-Resources Inv Rep 94-4223, 83 pp
DOI 10.1007/s10040-003-0294-0
