1 Introduction

The role of groundwater as a source of drinking water and other purposes for billions of urban or rural families cannot be ignored. Groundwater becomes a major source of water for fulfilling the various needs. According to a report on “Ensuring Drinking Water Security in Rural India” published by Department of Drinking Water and Sanitation, Ministry of Rural Development, Government of India (2009), 85% of the supply of drinking water is based on groundwater sources. Availability of drinking water during lean periods becomes a major issue. One of the major causes is that groundwater is over-extracted for industry and agriculture leading to depletion of drinking water sources.

Due to the depletion of groundwater, its management and conservation measures have become a challenging issue for the current scenario. These possible measures can be assessed by the development of various models for groundwater flow simulation which have also been used as decision support systems (DCS) (Bauer et al. 2005). The groundwater management includes efficient and effective management of groundwater resources at both quantity and quality level (Gorelick et al. 1984). The groundwater model is the replica of the actual groundwater system under certain conditions. It can be simple, if prepared on the basis of one-dimensional analytical solutions, or very complex and sophisticated, if prepared in a three-dimensional multi-layer form. Groundwater models can predict the future behaviour of the system if conditions remain the same or some alteration occurs in the condition (Olsthoorn 1985). For the development of efficient model, it is always recommended to start with a simple model, as long as the model concept satisfies modelling objectives, and then the model complexity can be increased step by step (Hill 2006). A good conceptual model should describe reality in a simple way that satisfies modelling objectives and management requirements (Bear and Verruijt 1987). In groundwater flow problems, boundary conditions are not only mathematical constraint, they also represent the sources and sinks within the system (Reilly and Harbaugh 2004). Selection of boundary conditions is critical for the development of accurate models (Franke et al. 1987). Omar et al. 2019) performed a modelling exercise using the analytic element method (AEM) and finite difference method (FDM) for the part of Ganga river basin.

The most widely used Groundwater flow model is MODFLOW, which solves the governing groundwater flow equations through Finite Difference Method (FDM) (Harbaugh and McDonald 1996). In literature, there are various studies based on MODFLOW such as Nishikawa (1998) Hunt et al. (1998), Ting et al. (1998), Olsthoorn (1999), Csoma (2001), Asghar et al. (2002), Faunt et al. (2004), Kumar & Kumar (2005), Umar et al. (2008), Igboekwe and Achi (2011), Khadri and Pande (2016). These studies are an example of the successful application of MODFLOW in the field of groundwater simulation. Previous regional modelling efforts include Prudicet et al. (1993), who simulated the regional flow in a large portion of the Great Basin with MODFLOW McDonald and Harbaugh (1988) using a cell size of 5 × 7.5 miles. The aquifer is conceptualized as a two-layer system, specified constant-head boundaries along with surface water features, and adopted constant-head and no-flow boundary conditions along with the remaining parts of the boundary. D’Agnese (1994) built a three-layer MODFLOW model to simulate regional flow in the Yucca Mountain region using a cell size of 1500 × 1500 m. Set of constant-head, no flow, and flux-specified boundaries were considered around the model due to the absence of real hydro-geologic boundaries. Igboekwe et al. (2008) studied groundwater flow characteristic near Kwa Ibo River in Abia state of Nigeria and developed groundwater flow model for assessing the degree of interaction between groundwater and Kwa Ibo River of the thick sandy aquifer. Zhou and Li (2011) discussed the large-scale regional groundwater flow model. Gaur et al. (2011) developed groundwater model for French aquifer. Kaviyarasan et al. (2013) developed a groundwater flow model for an unconfined coastal area surrounded by saline water bodies. Kumar Pradeep and Kumar Anil (2014) developed a groundwater model for Chouttupal Mandal, Nalgonda District Andhra Pradesh under steady state condition using FDM. Wang et al. (2009) also developed a 3D transient groundwater flow model for the Beijing Plain in order to analyze groundwater flow systems within the integrated regional water balance.

In this paper, study was carried out in and around Varanasi district, Uttar Pradesh, India to evaluate the groundwater resources and to understand flow dynamics in the region. The study proceeded with the conceptual model development of groundwater flow. The groundwater flow model is conceptualized by using information from a geological, climatic, and hydro-geological parameter of the concerned area. This model was prepared with five layers of the aquifer in the transient state condition to simulate the actual condition of the system. After model development, calibration of the model was done manually under the current stress conditions and calibrated model was used to evaluate the water stressed scenarios (Fig. 1).

Fig. 1
figure 1

Location Map of the study area

Full size image

2 The Study Area

The study area is located in the central Ganga plain of the Indian sub-continent. The study area consists of River Ganga in the south, River Gomti in the north and two small tributaries, namely Basuhi and Morwa in the eastern side of the area. The study area covers about 2785 km2 area in and around Varanasi district and lies between the latitude 25°05’ N − 25°40’ N and longitude 82°22’ E − 83°11’ E. The entire area is divided into 10 administrative blocks in which 8 blocks Pindara, Cholapur, Baragaon, Harhua, Chiraigaon, Sevapuri, Arjiline, Kashi Vidya Peeth exists in Varanasi District.

2.1 Physiography

The River Ganga is present as a trans-boundary between India and Bangladesh, the point of origin of the 2,525 km (1569 mi) long River is Western Himalaya in the Indian state of Uttarakhand, and it flows to the south and east Gangetic plains of North India into Bangladesh where it drains in the Bay of Bengal. Ganga is the longest River of India and is ranked second as the greatest River in terms of water discharge. Ganga basin is of magnificent variation in altitude, usage of the land, the pattern of crops, and climate. One of the tributaries of River Ganga is Gomti River which originates from the Gomat taal, Pilibhit India. The length of the river Gomti is around 900 km extending from the state of Uttar Pradesh and meets the Ganga near Saidpur, Kaithi in Ghazipur. One of the minor tributaries of the Ganga is River Varuna. Throughout the entire course, both the rivers receive a large amount of agricultural run-off consisting of pesticides and fertilizers from the catchment area.

2.2 Climate and Rainfall

The study area undergoes a humid subtropical climate with a large variation of temperature during summers and winters. The climate here is tropical with a marked monsoonal effect. The period between the middle of November to the end of February is the cold season. The hot season continues till the end of June. The rainy season spans over the period of mid June to September. The temperature in the study area during summer varies from 30 to 46 °C and in winter temperature varies from 9.5 to 16.5 °C. The average annual temperature is 26.1 °C. The study area receives about 80% of its annual rainfall i.e. 1,020 mm from the southwest monsoon during the month of July to August. In rainy seasons, the monthly rainfall variation is 96 mm to 290 mm whereas in the driest and wettest months, the difference in precipitation is up to 296 mm.

2.3 Aquifer Geometry

The study area forms a part of Vindhyan range comprising rocks of Proterozoic age. Fence diagram was prepared based on the lithological logs. Fence diagram reveals that the vertical and lateral disposition of aquifer exists up to the depth of 150 m The top layer of the study area is surface clay which varies from 3 to 6 m in depth. The second layer comprises of fine sand and varies spatially in nature. Below this layer, clay and sandy-clay layers exist. At the bottom, a relatively hard and impervious bed is encountered. Figure 2 shows the two sections of the study area. A conceptual model of five layers with thickness of 30 m each was used in the present study. Wells log data were used for development of aquifer geometry. After developing the stratigraphy for the study area, solid data was prepared, and interpolation of the solid data into the five-layered model was done.

Fig. 2
figure 2

Solid data cross section AA’ and BB’

Full size image

3 Methodology and Data

3.1 Groundwater Model Development

For the development of conceptual model, different layers were developed in vector data format. The different layer includes data of boundary conditions, elevation of the top and bottom layers, recharge rate, hydraulic conductivity, total water draft through pumping well, agriculture pattern, river characteristics, canal layout, surface water bodies etc.

For managing the draft through wells, data of water consumption for population, irrigation and livestock have been collected intensively by carrying out the field survey. Landuse classification of satellite imageries was also been done to calculate total agricultural area and other areas.

3.2 Borehole Module

Three-dimensional cross-sections were developed using several boreholes data. These Cross sections showed the soil stratigraphy data from the available drilling boreholes. These cross sections were later converted to solids which were used to characterize elevation data for the numerical model. Using borehole, module accuracy of the model was increased. In this study, collection of borehole data was done from U.P. Jal Nigam and created the 3D cross section. Using fifteen borehole data, fence diagram was created, which reveals the vertical and lateral position of aquifer, aquiclude, and aquitard in the study area down to a depth of 150 m BGL. The top layer of the study area consists of surface clay, which is varying in depth from 4 to 6 m bgl.

3.3 Boundary Conditions

3.3.1 River Boundary

The boundaries of the study area are considered as River boundaries and specified flow boundary for preparing the groundwater model (Fig. 3). The study area is surrounded by River Gomti in North, River Ganga in the south, Varuna in the east along with two tributaries Morwa and Basuhi. Intensive river survey was carried out on the River Ganga, Varuna, Morwa, Basuhi, and Gomti to measure the stages and cross sections of the Rivers. The survey was conducted at different locations for the year 2015, 2016 and 2017. For the year of 2006 and onwards, data were collected from the Central Water Commission (CWC), Varanasi. The data were used as input in the model as River boundary. River bed conductance values ranging from 30 to 150 m2/d were adopted for the different Rivers (Maheswaran et al. 2016).

Fig. 3
figure 3

Topographic Elevations of the study area with River boundary condition, and Grid used in the modeling

Full size image

3.3.2 Specified flux Boundary

Two boundaries of the study area were taken as specified flux boundaries where one boundary is situated between Gomati River and Basuhi River and another boundary is situated between Varuna River and Ganga River. The length of specified flux boundaries is 16.72 and 18.56 km, respectively. The specified flux from the contributing regions was estimated through groundwater contour data.

3.4 Top and bottom layers

The topography of the study area is largely even with the low relief and less undulation. Realistic representation of the top and bottom surfaces in the model is a difficult task, and the accuracy of the model result depends on these surfaces. The top surface of the model was determined by DEM of the study area. In this study, Shuttle Radar Topography Mission (SRTM) DEM was used, which was taken from the USGS site (https://earthexplorer.usgs.gov). The bottom layer was defined with the help of well log data collected from Uttar Pradesh Jal Nigam. Evenly distributed data over the entire area were taken and interpolated to generate the bottom surface.

3.5 Depth of Groundwater Level

The groundwater level in the study area is obtained from site India-WRIS (Water Resource Information System of India) portal (http://www.india-wris.nrsc.gov.in). From India-WRIS, the groundwater level is obtained at different location within the study area. In pre-monsoon seasons i.e., June 2015 and June 2016 the depth to water level ranges from 2.5 to 15 m and 1.8-13.48 m below ground level, respectively. The deeper water level viz. 11.03, 13.48, and 15.0 m were recorded at Pali, Rustampur, and Thatra, block respectively. Similarly, in monsoon season i.e., June 2015 and June 2016 range of water level depth is 2.23-19 m and 2.34–20.1 m, respectively. In post-monsoon season i.e., November 2015 to November 2016 depth of water level is in range of 2.36–18.2 m and 2.78–19.45 m respectively.

3.6 Recharge

Major groundwater recharge in the study area is through Rainfall, Canal seepage, and Irrigation return flow. The recharge enters the aquifer through infiltration in the overlying confining unit and seepage through riverbeds. Study area receives about 80% of its annual rainfall, i.e., 1020 mm from south-west monsoon during the period of July to August. Rainfall data were procured from Central Water Commission (CWC), Varanasi for the period of 10 years from 2006 to 2016. Canal seepage was assumed as 10% of the water released for irrigation. The value of canal seepage depends on the type of canal, i.e., lined or unlined (Rao 1993).

Land-use and land-cover pattern was obtained through satellite image classification. Images were collected once in four years, which were then taken the same for the subsequent four years. The image was classified for the year of 2006 and then used for the period of 2006–2008. Further satellite imagery for the year of 2009 was classified, which was further used for the period of 2009–2012 and so on. After the classification of images, different recharge zones were marked. Every recharge zone have different recharge coefficient depending on the type of soil available, and the amount of rainfall occuring in that recharge zone. There are empirical equations that can be used in the groundwater recharge estimation. These empirical equations relate the rainfall with the groundwater recharge. A modified version of Chaturvedi’s (1973) equation was proposed by Irrigation Research Institute, Roorkee, which was used in this study.

$$\text{R}={1.35 (\text{P}-14)}^{0.4}$$

Where R is the net recharge due to rainfall and P is the rainfall measured in inch. Recharge can be converted into millimetres (mm).

The recharge coefficient can be defined as the ratio of recharge to effective rainfall, expressed in % (Muteraja 1986).

$${\text{R}}_{\text{c}\text{o}\text{e}\text{f}\text{f}}=\left(\raisebox{1ex}{$\text{R}$}\!\left/ \!\raisebox{-1ex}{${\text{P}}_{\text{e}\text{f}\text{f}}$}\right.\right)\text{\%}$$

Where R is the net recharge and Peff is the effective rainfall.

3.7 Hydraulic Conductivity

The study area is a part of Indo-Gangetic plain which is underlain by quaternary alluvial sediments of pleistocene to recent times. However, in the study area, unconsolidated sediments form a sequence of clay, silt, and sands of various grades. The presence of kankar carbonates at different location is intercalated with the clays and sands, which forms the potential aquifers at various depths. The top layer throughout the Ganga plain consists of clay layer for a few meters. The total thickness of potential aquifer varies upto 150 m. Hydraulic conductivity was computed for different locations, based on the test samples. 25 samples were collected and K values were computed correspondingly. The value of hydraulic conductivity varied from 12 m/day to 24 m/day.

3.8 Well Discharge

Water demand for the irrigation purpose is fulfilled by wells and in some places by canals as well. Whereas, domestic water demand is fulfilled by the wells in the study area. Water demand for agriculture was calculated by the classification of satellite images and then calculating the area covered by agricultural field. Further classification for each type of crop was done and corresponding area was calculated for the each crop. The imageries were classified in five groups i.e. Vegetation, Urban/Built-Up land, Water, Sand/waste Land, Agricultural field. Table 1 shows the blockwise LU & LC details for the whole study area. Intensive field survey was done throughout the study area to know the agriculture pattern from the natives and field truthing of the satellite imageries. Based on the type of crops that were harvested, the water requirement for each crop was calculated for per meter square area. As the area irrigated by the particular crop is known, water requirement for the entire area for that crop can be calculated. In this fashion, water requirement for agriculture for the entire study area was calculated.

Table 1 Value of hydraulic conductivity (m/day) at 25 different locations
Full size table

There are ‘n’ numbers of crops that are harvested in the study area. Where “w” denotes water requirement of each crop planted in per m2 and the different crop is planted in an area of “A.” Therefore the total water required for agriculture in the study area was calculated as:

$${\varvec{w}}_{\varvec{A}}= \sum _{\varvec{i}=1}^{\varvec{n}}{\varvec{w}}_{\varvec{i}}{\varvec{A}}_{\varvec{i}}$$

Where ‘i’ is the no. of crops planted in the study area, ‘WA’ is the total water required by agriculture. Total water required for agriculture was found to be 225.86 Mm3 per month. After calculating the total agricultural area, individual area for each type of crops was calculated. This exercise was done by field survey and discussion with the farmers of different villages to understand yearly cropping pattern in the area. The major crops in the area were found as rice, wheat, and pulses, vegetables. Then the agricultural area of a particular type of crop was multiplied by per m2 water requirement of that crop, according to the season, to estimate the total water requirement. In the study area, three type of crops were found. The duty of rice, wheat, and pulses were taken 1250 mm, 550 mm and 350 mm, respectively (Garg 2016) (Tables 2 and 3).

Table 2 Detail of land use/land cover of the study area
Full size table
Table 3 Agricultural groundwater demands in the study area
Full size table

Water requirement for domestic demand was calculated by assuming per capita demand of 135 L/person and 170 L/person for the rural and urban population respectively (Modi 1998). Data related to population of the study area were taken from Census of India. In this way, water demand of the study area was calculated.

$$W_D= \mathit {135}\ ^\star{P_r}\ or\ W_D= \mathit {170}\ ^\star{P_u}$$

Where ‘WD’ is the total water demand in the study area,‘Pr’ is the total population in the rural area and ‘Pu’ is the total population in the urban area. Total water required for domestic demand was found to be 15.13 Mm3 per month. Table 4 shows the distribution of water demand in each administrative block.

Table 4 Livestock population in the study area
Full size table

Different livestock consume different amount of water for fulfilling their thrust. Considering their average water demand, total water demand for livestock at blocks level was calculated. The average drinking water demand of cow, buffalo, goat, pig, sheep and poultry were taken as 30, 35, 3, 3.5, and 1 L/day respectively (http://upenvis.nic.in). The Table 5 shows groundwater demand for the livestock. Total water demand for livestock was calculated as 0.253 Mm3 per month.

Table 5 Livestock groundwater demand in study area
Full size table

Thus, the combination of the above demands is comprised of the total water demand for the study area. Total water demand was used to distribute the water demand in the pumping wells of the study area. Specific yield of the aquifer in the study area was found between the range of 20 lps to 25 lps (1728 m3/day). Based on the specific yield of the aquifer, the maximum well capacity per day was calculated for the study area. To get the number of pumping wells in the study area, the total water demand was divided by per day well capacity.

3.9 Observation Wells

The groundwater level at different location in the study area was taken from India-WRIS website (www.india-wris.nrsc.gov.in). The water table data was taken for four different seasons of the year namely: pre-monsoon, post-monsoon rabi, monsoon, post-monsoon kharif for different evenly distribute wells. Observed groundwater head from the observation wells was taken from the CGWB, Varanasi for four periods; pre-monsoon ravi (January-March), pre monsoon (April-June), monsoon (July-September) and post monsoon kharif (October-December) was compared with the simulated head.

4 Groundwater Model Conceptualization

The groundwater flow model was conceptualized on the basis of geological, climatic, and hydro-geological characteristics of the study area. Thus the GIS based conceptual model groundwater was developed using MODFLOW based software GMS 10.2 (Aquaveo 2010). MODFLOW can solve the steady and transient groundwater flow equation and can simulate single or multiple aquifer layers with different boundary conditions. (Tamma Rao et al. 2012; Surinaidu et al. 2013; Senthilkumar and Elango 2004).

The area of the aquifer is about 2785 km2, and the thickness of the aquifer varies up to 150 m. Transient state simulation of groundwater model was done using five layers in the computational grid. The grid consists of 210 rows and 210 columns where each cell measures 250 m × 250 m on ground. For transient condition, ten years data was taken from the Central Ground Water Board (CGWB), Varanasi for January 2006 to December 2017. Transient data was entered in the conceptual model using time values where the time at the beginning of the first MODFLOW stress period was considered as the reference time. In the model, data were entered to the model at the interval of 91 days, beginning from January 2006, i.e. pre monsoon Ravi up to December 2017 i.e. post monsoon Kharif. Thus, there were total 48 stress periods for which head stage, bottom elevation for rivers, specified flow, and recharge rates were entered. Figures 4 and 5 shows the variation in the river head stage and bottom elevation of the river Ganga, with different time duration starting from January 2006 to December 2017.

Fig. 4
figure 4

Variation of the Ganga River head stage

Full size image
Fig. 5
figure 5

Variation of the Ganga River bottom elevation

Full size image

The boundary condition of the aquifer was considered as time invariant. Gomti River in the north, Ganga River in the south, Basuhi and Morwa River in the east of study area was considered as river boundary. The portion between Gomati and Basuhi River; and between Ganga and Varuna River was taken as specified flow boundary. The specific flow entering the study area, from Basuhi River, was taken as 0.000005 m3/day and the specific flow, from Varuna River, entering the study area was taken as 0.0000038 m3/day which we were computed with the help of groundwater table of the area and further calibrated.

The starting head at the inlet of the model boundary was taken as 66.279 m, whereas at the downstream end, constant head was taken as 60.78 m. Starting constant head at the Gomti River entering the study area was taken as 68.23 m, and at the downstream end, constant head was taken as 60.78 m. Starting constant head at the Morwa River and Basuhi River were taken as 73.17 and 69.83 m, respectively. All input data were provided in the model as a vector layer (Table 6).

Table 6 Total population and groundwater demand in the study area
Full size table

4.1 Model Output

Groundwater flow model provides the computed groundwater head. A graph between the observed head and the computed head was made to identify the error in the model. Different graphs were generated such as residual error v/s observed data, error v/s time step and error v/s simulation, Time series also gives a better understanding of the behaviour of the model.

4.2 Model Calibration

Model calibration process is required to minimize the difference between model output with real field conditions up to acceptable limit. In this process, values of different input parameters are varied so that the field conditions at a site should be properly characterized. In this study, the model was calibrated for transient state conditions in which simulations involved the change in the head (groundwater head and river stage) with time. Initial values of hydraulic conductivities and recharge rate were introduced into the model. Further, residual error in the initial model solution was analyzed, and a new value for hydraulic conductivities and recharge rates were entered. After running the groundwater flow model, a new MODFLOW solution was generated, and again, a new error estimate was computed and plotted. This process was repeated until the error was reasonably small.

Fourteen observation wells were taken in the study area. For each observation well, the observed head data was categorized at 180 days interval, and model calibration was done at the same time interval with 90% confidence level. This confidence value represents confidence in the error estimate.

Residual error in the initial model solution (run 1) was analyzed. In run 2, a new value for hydraulic conductivity was entered by increasing this value by 10%, and recharge rate was decreased to 0.0004 m/d. After running the groundwater flow model, a new MODFLOW solution was generated, and again, a new error estimate was computed and plotted. This run reduced the difference between the observed head and simulated head. In run 3, the conductivity was further increased by 20%, and recharge rate was decreased to 0.0003 m/d. This run again, reduced the root mean square error (RMSE). In run 4, the hydraulic conductivity value was increased by 15%, and recharge rate was further decreased to 0.0002 m/d. This run provided a better match as it gave very less standard error of estimate to 0.002 m. The good calibration shows that this model is actually representing the real field conditions for the study area. The parameter values used in calibration of the model are listed in Table 7. Figure 6 represents the error in head value. Green colour bar shows that computed value lies entirely within the target value, yellow colour bar shows that compute value is outside the target but error is less than 200% and red colour bar shows error in the computed value is more that 200%. For this calibrated model, all colour bars lies in the target limit, this represents that this model is well calibrated. Time series graph is presented in Figs. 7, 8 and 9. This graph was prepared for three observation wells, namely Varanasi, Barwaon, and Aurai. In each graph, head comparison was done for different stress period. Figure 10 shows the calibration target with the observed head values.

Fig. 6
figure 6

Groundwater flow lines along with colour bar

Full size image
Table 7 Parameters values used in the model calibration
Full size table
Fig. 7
figure 7

Time series graph for Varanasi Observation well

Full size image
Fig. 8
figure 8

Time series graph for Barwaon Observation well

Full size image
Fig. 9
figure 9

Time series graph for Aurai Observation well

Full size image
Fig. 10
figure 10

calibration targets with the observed head values

Full size image

Scatter plot was drawn, as shown in Fig. 11 (observed and compute head). Scatter plots show calibration targets versus simulated values and allow for a quick assessment of model fit. Scatter plots also visualize bias in the calibration. Bias is absent when points in the scatter plot are more or less equally distributed around the central line shown on the plot, which indicates one-to-one correspondence between simulated and observed values. The calibrated model was validated for the period of January 2014 to December 2017. The observed root mean square error (RMSE) was found within the allowable limit i.e. 1 m. Fig. 11 shows a good match between the observed and computed groundwater heads.

Fig. 11
figure 11

compute and observed head for fourteen observation wells

Full size image

5 Result and Discussion

The average groundwater depth in the basin was found 6 m BGL in pre-monsoon and 10 m BGL in the post-monsoon period. In and around the Kashi Vidya Peeth block of the study area, spatially distributed groundwater level ranges from 8 to 15 m bgl which shows the highest drawdown along with Chiraigaon block shows maximum drawdown at some places. Area covered with vegetation like Banaras Hindu University (BHU) and cantonment region, groundwater level ranges from 1.5 to 6 m BGL. Deep water level of more than 18 m BGL occurred in the north-east part of the study area and shallow groundwater level, i.e. less than 4 m BGL found near to the river Ganga, Varuna, and confluence of Gomati and Ganga River.

From the model, it was observed the groundwater flow direction was towards the rivers, as shown in Fig. 12. In the figure, circled C part showed that there was a movement of water from the aquifer to river side and circled A and B indicated that there was outward movement of flow from the study area.

Fig. 12
figure 12

Groundwater directions for the study area

Full size image

5.1 Scenario Generation

The results show that the groundwater level in the Varanasi region is falling due to over-exploitation of the water resources. As Varanasi is a tourist place (floating population is approximately 25,000 per day) (JNNURM Final Report, 2006). Recently, the Indian government has approved many projects under a master plan for 2031 and smart Kashi, a smart city proposal. Considering these changes for the study area, three different scenarios were generated, and the model was simulated for those scenarios.

  • Scenario 1: Keeping in mind, the population of the study area for 2031, discharge rate (pumping rate) was increased.

  • Scenario 2: Considering the urbanization for 2031, the recharge rate was decreased to 0.00008 m/d.

  • Scenario 3: Taking above both Scenarios simultaneously.

The observed and simulated groundwater levels for the three scenarios were studied For the Scenario 1, the population was predicted for the year 2031 and based on that groundwater demand was calculated as 25.193 Mm3 per month. To compensate these demands, pumping wells were increased in the study area and model was simulated with increased discharge values. Results show that the drawdown was very uneven. At some places, it decreased more i.e. upto 6 m as compared to another prominent place where the chance of depletion of groundwater was more i.e. upto 4 m. Groundwater level was found less decreasing near the rivers due to the transfer of water from the river to the aquifer as additional groundwater discharge increases river water seepage into the adjoin aquifer. This shows that a gradual rise in the groundwater pumping can increase river seepage to the aquifer.

When considering scenario 2, the lower recharge rate was taken to accommodate the decrease rainfall distribution due to climate change. Discharge rate was taken as same. Results show less change in the groundwater level as compare to Scenario-1.

In the third scenario, both changes were taken simultaneously. The results show that discharge has sufficiently more domination over the recharge rate. Additional groundwater pumping can lead to change to flow direction from the aquifer to river and vice versa.

6 Conclusion

Groundwater modelling is very important tool to understand the dynamics of groundwater behaviour, and correspondingly can be used as to identify the best management practices for groundwater resources conservation. Modelling of the groundwater is a complex phenomenon because various hydro-geographical features are required to find the solution. Nowadays, the groundwater model has proven to be a useful tool over several decades for addressing a range of groundwater problems and supporting the decision-making for the better management of the groundwater. In this study, transient groundwater modeling was done for the middle part of the lower Ganga basin near Varanasi and adjoining areas. The model was developed using MODFLOW. Model calibration was also performed by selecting three model parameters.

The results of the model show that the computed values of the water head are in good-fitness of the measured data, which indicate the model is reliable. Collection of model input data is very challenging task and can be generated through different sources and techniques particularly in the absence of direct field values ex. discharge from each wells. Results show that the explicitly computed well discharge data can be used for model development. The field monitoring was also incorporated in the study to verify the model predictions. Similar studies can be performed for other water-stressed areas for reliable water resources estimation so that better and efficient water resources planning and management can be done.