Assessing the Impact of Pattiseema Lift Irrigation Project on Groundwater Resources in West Godavari District Using Visual MODFLOW

Uday Bhanu Prakash Vaddevolu^1,2, Sasikala D³, Vivek Sharma^1,*

Published in Journal of Natural Resources and Agricultural Ecosystems 3(1): 17-35 (doi: 10.13031/jnrae.16178). Copyright 2025 American Society of Agricultural and Biological Engineers.

¹ Agricultural and Biological Engineering Department, University of Florida, Gainesville, Florida, USA.

² Biological and Agricultural Engineering Department, Texas A&M University, College Station , Texas, USA.

³ Kerala Agricultural University, Vellanikkara, Kerala, India.

^* Correspondence: vsharma1@ufl.edu

The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License https://creative?commons.org/licenses/by-nc-nd/4.0/

Submitted for review on 2 September 2024 as manuscript number NRES 16178; approved for publication as a Research Article by Associate Editor Dr. Meetpal Kukal and Community Editor Dr. Kati Migliaccio of the Natural Resources & Environmental Systems Community of ASABE on 15 January 2025.

Citation: Vaddevolu, U. B. P., D, S., & Sharma, V. (2025). Assessing the impact of Pattiseema lift irrigation project on groundwater resources in West Godavari District using visual MODFLOW. J. Nat. Resour. Agric. Ecosyst., 3(1), 17-35. https://doi.org/10.13031/jnrae.16178

Highlights

• Impact of Pattiseema Irrigation on groundwater recharge is evaluated in West Godavari District.
• Spatial variation showed significant groundwater rise, highest at 53% in high-land region.
• Potential groundwater zones were identified in the study area.
• Groundwater rose by 3 m near the unlined canal, indicating a positive impact on water resources.

Abstract. Changes in groundwater recharge processes caused by regional changes in water use patterns need to be fully considered for the accurate budgeting and management of water resources. The West Godavari District (WGD) of India plays a significant role, benefiting from abundant water resources and bounded by the Godavari and Krishna rivers; the former discharges around 85,000 million cubic meters into the Bay of Bengal, while the latter has been susceptible to recurring droughts. To address this, the Pattiseema lift irrigation project was initiated, diverting surplus water from the Godavari to the Krishna River via unlined canals, significantly contributing to groundwater recharge. This study investigates the spatial and temporal variation of groundwater resources in the WGD and identifies the potential groundwater zones. In addition, groundwater flow simulation was performed using the Visual MODFLOW under steady and transient state, to assess the impact of the Pattiseema lift irrigation project on these resources. To achieve this, the study region was divided into three regions, i.e., low-land, mid-land, and high-land regions. Variation in groundwater levels was observed between pre- and post-monsoon periods, with groundwater levels rising to 35%, 39%, and 53% post-monsoon in low-, mid-, and high-land regions, respectively. This variation is due to the strong spatial pattern of annual precipitation across the study area. There was an approximate difference of 262 mm average annual precipitation between northeast and southwest. This variation can have an influence on the groundwater recharge and availability. Groundwater Potential Zones (GWPZs) were identified, with 740, 436, and 47 villages located within the medium, moderately high to high, and low GWPZs, respectively. The Visual MODFLOW resulted in a satisfactory agreement between observed and measured water heads with an R² of 0.89, an MSE of 1.1 m, an RMSE of 8.14 m, and an NRMSE of 6.47%. The simulated model revealed the effect of the unlined canal on the groundwater table in the study area, showcasing a rise of approximately 3 meters in the study area. This observation signifies that the unlined canal has a positive impact, enhancing groundwater resources in proximity. This outcome emphasizes the critical importance of well-planned water management initiatives, emphasizing their role in sustaining and improving local hydrogeological conditions.

Keywords.Groundwater, India, Irrigation, Modeling, Precipitation.

Groundwater represents the largest freshwater resource in dry and semi-arid regions of the world, including India. Globally, groundwater resources provide approximately 50% of drinking water, 40% for agricultural use, and 15% for industrial applications (Awais et al., 2021; Sharma and Bhambota, 2022). In India, groundwater has traditionally been the primary source to meet the domestic water needs of over 80% of India’s rural population and 50% of its urban population (Adimalla and Qian, 2021). Moreover, about 70% of the country’s irrigated food production relies on well irrigation (Panneerselvam et al., 2023). Agriculture, being the cornerstone of the Indian economy, occupies approximately 60.4% of the total land area. While rivers have historically served as a major source of irrigation, they have undergone significant changes due to urbanization and climate change, resulting in reduced agricultural productivity. While some areas are experiencing severe floods, others are affected by droughts. This situation has led to a drastic increase in the utilization of groundwater as an alternative to address water scarcity while focusing on increasing the groundwater recharge rate. Several rivers in the southern part of India are on the verge of drying up due to consecutive droughts, while a few rivers are having surplus water, which ultimately drains into the sea. Addressing this challenge requires effective management of both surface and groundwater resources. Notably, groundwater has become a predominant source, especially in areas where surface water availability is declining (Mondal et al., 2010; Suhag, 2016).

Groundwater level mainly depends on the depth of water-bearing strata from the ground surface and soil properties. However, the significance of groundwater is heightened when subsurface formations are comprised of hard-rock aquifers. The potential for groundwater in semi-arid hard-rock environments is primarily restricted to shallow areas that have undergone weathering and fracturing. Unlined canals have been recognized as significant contributors to groundwater recharge, particularly in regions with permeable soils and extensive canal networks. For example, in the lower Bhavani Basin, unlined canals facilitate groundwater recharge through seepage losses, which are a critical source of subsurface water replenishment (Mirudhula, 2014). Similarly, in the Upper Ganga Canal command area, seepage from unlined canals was reported to enhance natural groundwater recharge, supporting both agricultural and domestic water needs in the region (Kumar and Seethapathi, 2002). In southern India, unlined canals have been identified as vital groundwater recharge structures, particularly in drought-prone areas where declining water tables are a pressing concern (Shah, 2020). Singh (2022) highlighted that while lining canals reduces water losses and improves delivery efficiency. This dual role of unlined canals highlights the need to balance surface water conservation with groundwater recharge to ensure sustainable water resource management in areas affected by climate variability. However, the indiscriminate exploitation of shallow groundwater resources has been undertaken in response to the ever-growing demands for freshwater. For example, a census report published by the Government of India indicated that the median percent decline in shallow well irrigated area from 1996–2000 to 2006–2010, across the South Indian states of Andhra Pradesh, Telangana, and Karnataka, is 52%, while the median percent increase in deep well irrigated area over these states is 98% over the same time frame (Hora et al., 2019). The issue pertaining to the declining availability of groundwater in southern India and the potential implications of future climate change on an already critical state of affairs have been widely acknowledged in scholarly discourse (Hora et al., 2019). Similarly, in northwestern states, it is observed that groundwater is declining by around 3 cm per year from 2002 to 2008 (Sarkar et al., 2020). This represents the critical condition of groundwater availability. In this situation, optimal use and artificial recharge of groundwater are the alternatives to overcome water scarcity (Shankar et al., 2011). Hence, it is imperative to forecast the dynamics of groundwater levels in hard-rock aquifers at both spatial and temporal scales in order to develop effective strategies for the optimal planning and management of limited groundwater resources.

Groundwater models play an important role in the development and management of groundwater resources and in predicting the effects of management measures. The monitoring of groundwater levels is essential for obtaining discrete data in both spatial and temporal dimensions (Meggiorin et al., 2024; Siade et al., 2017; Keery et al., 2007). The dynamics of groundwater levels are influenced by two primary factors: groundwater recharge and subsurface flow, in addition to the properties of the aquifer and soil. Subsurface flow is predominantly governed by the hydraulic gradient and conductivity of the aquifer. These measurements are necessary for conducting a quantitative evaluation of groundwater resources. Various conceptual, mathematical, physical, and digital modeling approaches coupled with geostatistical techniques have proven to be valuable tools in the examination of intrinsic uncertainties within groundwater systems (Venkatramanan et al., 2019). Visual numerical modeling techniques such as the Finite Element subsurface FLOW system (FEFLOW) (Diersch, 2005), Visual Modular Three-Dimensional Flow (Visual MODFLOW) (Waterloo Hydrogeologic, 2000), a 2D and 3D geo-statistics, uncertainty analysis, and visualization software package (UNCERT) (Wingle et al., 1999), and Processing MODFLOW for Window (PMWIN) (Chiang, 2005) have been extensively used to monitor the dynamics of groundwater and to predict future acquire conditions applied in different regions of the world. Among various groundwater models, MODFLOW (McDonald and Harbaugh, 1988) was used in many groundwater applications because of its simplicity, modular program structure, and separate package to resolve special hydrogeologic problems. These techniques can be effectively applied to address various groundwater estimate challenges, such as interpolation, integration, and differentiation (Patel et al., 2023). For example, Nandimandalam et al. (2023) developed a groundwater model for the Nellore coastal zone basin to predict the potential environmental and socioeconomic impacts of the groundwater abstractions. They estimated the aquifer parameters by developing a transient groundwater flow model in Visual MODFLOW and predicted the groundwater head for future years in the Nellore district situated at the seacoast. Kumar and Kumar (2014) developed a steady-state MODFLOW model to quantify groundwater in Choutuppal Mandal, Nalgonda, AP, using groundwater data from 19 observation wells. The model was conceptualized as a two-layered, weathered, and fractured aquifer system spread over a 19215 m x 10366 m area and observed and computed groundwater level contours were found in good agreement. Swathy (2016) developed a MODFLOW model for groundwater resource flow modeling and mapping of the Kadalundi River basin of the Malappuram district of Kerala. The model was developed to study the aquifer characteristics and spatial and temporal groundwater fluctuations in the area. The model was calibrated for steady state and transient conditions and predicted the flow head and groundwater conditions for various pumping rates and recharge. Predictions were made for the next 5, 10, and 15 years with a 5% decrease in recharge for every year and predict the groundwater condition after 15 years by increasing the pumping rate by 10%, 25%, and 50% of the pumping rate of the validated period.

The West Godavari District (WGD) of India plays a significant role, benefiting from abundant water resources and bounded by the Godavari and Krishna rivers; the former discharges around 85,000 million cubic meters into the Bay of Bengal, while the latter has been susceptible to recurring droughts. Anticipating the challenges posed by climate change and the need to conserve water resources, the Government of Andhra Pradesh decided to construct Pattiseema lift irrigation to interlink the Godavari and Krishna rivers. This lift irrigation project is used to divert the surplus water from Godavari River to Krishna River. Since the groundwater potential of the area is not adequately investigated, lack of knowledge is a problem for the development and management of the aquifer in the area. To ensure judicious management of groundwater, the present status should be evaluated, and a prediction of the future water resource scenario should be attempted. This can ensure proper groundwater management of the area. So, the specific objectives are to (1) evaluate the spatial and temporal variation of precipitation and pre-monsoon and post-monsoon seasons groundwater levels in the West Godavari district (WGD) and (2) test the performance of the MODFLOW model to evaluate the influence of a lift irrigation project on this groundwater recharge in WGD.

Materials and Methods

Study Area

The study was conducted to investigate the groundwater resources in WGD, which lies between the latitude of 16° 15’N to 17° 30’N and longitude of 80° 50’ E to 81° 55’E in Andhra Pradesh state, India. Situated in the western delta and occupying the total area of 850,665 ha (fig. 1). The WGD comprises 46 mandals, which include 910 villages, and is further divided into 8 municipalities, covering an area of 7,795 square kilometers. The district experiences a wide range of seasonal variation in temperature and precipitation. In general, the study area experiences a normal annual rainfall ranging from 840 to 1283 mm between 2004 and 2017, with an average of 1078 mm. The southwest monsoon contributes to a major part of rainfall, followed by the northeast monsoon. The average minimum temperature in the area is approximately 23.2°C, while the average maximum temperature is around 35°C. Likewise, the relative humidity fluctuates between 62% and 77%.

Figure 1. Location of the study area along with the elevation, drainage density, and slope map of the West Godavari district.

The WGD is bounded by the Godavari River in the East, the Krishna River (which is 38.62 km in the West), and the Bay of Bengal in the South. The study area is mainly drained by the Godavari, Yerrakalva, Tammileru, and Ramileru rivers. The Godavari is the major river flowing in the region, forming the large-scale deltaic plains, where the prevailing soil type is predominantly sandstone, followed in frequency by clay, sand, clayey sandstone, granite gneisses, and shales. Other rivers in the district join Kolleru Lake and it is located in the southern part of the district. Kolleru Lake is India’s biggest freshwater lake, and its drainage seems to be dendritic. The study area comprises of different land-use patterns, including land, forests, barren and uncultivable land, cultivable waste, permanent pastureland, fallow lands, and fish and prawn culture. The primary crops cultivated in this region are rice (Oryza sativa), jowar (Sorghum bicolor), and maize (Zea mays), followed by finger millet (Eleusine coracana), cashews (Anacardium occidentale), sugarcane (Saccharum officinarum), coconut (Cocos nucifera), and various vegetables.

The state’s surface and groundwater resources are regulated by the government of Andhra Pradesh as per the Water, Land, and Tree Act of 2002 (WALTA). Surface water is a major source for irrigation in the study area, whereas groundwater is the source in upland regions, where canal water is not available for irrigation. Many major, medium, and minor irrigation projects were initiated by the state government to recharge the groundwater in the study area, including (1) the Tammileru Reservoir Project, covering an area of 3,712 ha; (2) Vijayarai Anicut, covering an area of 4,340 ha; (3) Jalleru Reservoir Project, covering an area of 1,700 ha; and (4) Yerrakalava Reservoir Project, covering an area of 13,709 ha (Directorate of Economics and Statistics, 2015). To further enhance the recharge capacity of WGD, the Pattiseema lift irrigation project was completed by the government of Andhra Pradesh in 2017 by connecting the Godavari and Krishna rivers through an unlined canal. The primary purpose was to redirect the excess water from the Godavari River to the Krishna River, which would potentially lead to groundwater recharge near the unlined canals. Land use and land cover details of the WGD are shown in table 1 (Directorate of Economics and Statistics, 2015).

Data Collection and Identification of Potential Groundwater Zones

This study was focused on evaluating groundwater levels throughout the specified region to identify areas with potential for groundwater extraction. The groundwater data from unconfined and confined aquifers from 55 observation wells and 70 piezometers were collected from the Groundwater and Water Audit Department, Government of Andhra Pradesh, from 2003 to 2018. These datasets were reported at an interval of two months. For some piezometers, head data was missing or incomplete from 2003 to 2018. Therefore, to understand the spatial and temporal variation of groundwater levels, only observation well data from 55 monitoring wells were used in our analysis. The data associated with groundwater recharge and subsurface flow, along with properties of the aquifer and soil associated with each observation well and piezometer, were obtained from the Groundwater and Water Audit Department, Government of Andhra Pradesh. Point coverage of observation wells was created in ArcGIS. The location (longitude and latitude) of the observation wells and the water level data were imported into a geodatabase and explored using the ArcGIS (ESRI, Redlands, Cal.) Geospatial Analysis Tool before interpolation (Baghel et al., 2023).

Regional differences in the environmental characteristics, with the combined effect of climatic conditions and soil and topographic characteristics, divide the WGD into three broad topographical regions: low-land (elevation up to 7.5 m), mid-land (elevation between 7.5 and 75 m), and high-land (elevation exceeding 75 m), as shown in figure 1. Among the 55 observation wells in the villages, 5 are situated in low-land, 22 in mid-land, and 28 in high-land. The spatial variability of the groundwater across the study area was investigated by comparing the pre-monsoon (May) and post-monsoon season (November) for the years 2013, 2015, and 2017. Additionally, temporal variation in groundwater levels was studied by observing the groundwater level trends over a period of 14 years, from 2003 to 2017. The predicted values of groundwater level data based on 14 years of historical data were computed using the kriging interpolation method. Kriging is most appropriate when it is known that there is a spatially correlated distance or directional bias in the data (Sharma and Irmak, 2012; Sharma et al., 2013; Merino et al., 2001).

Groundwater Modeling

In this study, groundwater flow simulation was performed using the Visual MODFLOW (McDonald and Harbaugh, 1988) in WGD to assess the impact of the Pattiseema lift irrigation project on these resources. MODFLOW has been widely adopted as a calculation program due to its user-friendly, modular program structure, and separate packages designed to address specific hydrogeologic challenges. Notably, its seamless integration with Geographic Information Systems (GIS) eases the incorporation of spatial data, enhancing model parameterization accuracy. The flowchart below outlines the key stages in the model development process, as shown in figure 2.

The robust numerical solver within MODFLOW enables the simulation of intricate hydrogeological conditions, including transient flow, heterogeneous aquifer properties, and variable boundary conditions (Akhter and Hossain, 2017; Chowdhury and Rahnuma., 2023). The Integrated Finite Difference Method (IFDM) serves as the simulation engine, assuming constant density and viscosity of groundwater under transit state conditions, adopting a conceptually straightforward mathematical approach (McDonald and Harbaugh, 1988). This involves transforming various equations to simulate groundwater flow in the three-dimensional direction of the study area as:

(1)

where

K_x, K_y, and K_z = hydraulic conductivity in three different directions

S_s = specific storage

h = hydraulic head

Since the hydraulic head value does not change with time under steady-state conditions, which is expressed as:

(2)

The visual MODFLOW has several solver packages to solve the numerical modeling equations for groundwater flow simulation, including the Preconditioned Conjugate Gradient (PCG), Strongly Implicit Procedure (SIP) package, WHS solver for Visual MODFLOW package (WHS), Slice Successive Over Relaxation (SOR) package, and Geometric Multigrid solver (GMG). In this study, the Strongly Implicit Procedure (SIP) package was used to resolve the partial differential equations through iterative procedures.

Figure 2. Methodological flowchart for groundwater modeling.

The model domain was established by incorporating a base map, specifying 125 rows and 150 columns. Areas other than the study area were excluded by inactivating the cells in the base map in order to focus solely on the study area. Elevation was extracted using the Digital Elevation Model (DEM) using ArcMAP. This data was subsequently imported and assigned to the model domain, along with the corresponding latitude and longitude information. Following that, each soil layer was imported in accordance with the identified soil profiles. Sandstone, clay, sand, clayey sandstone, and granite gneisses are the major soil types found in the study area. Soil profile data and aquifer properties like hydraulic conductivity, storage capacity, and specific storage data were also collected from the Department of Groundwater and Water audit in Andhra Pradesh, India. This meticulous process was employed to ensure that the model's x, y, and z coordinates closely matched those observed in the field study, thereby enhancing the precision of the simulation.

After creating the model domain, observation wells, piezometers, and pumping wells data were assigned in the model, which plays a pivotal role in characterizing groundwater levels and discharge in the study area. Furthermore, pumping well data, indicating either groundwater extraction (+) or recharge (-), was seamlessly integrated into the model domain. Further, aquifer properties were assigned in the model development process to understand the water movement through the soil profile. Hydraulic conductivity (K_x, K_y, K_z), specific storage, and specific yield of different soil layers were assigned to each layer in the model. Hydraulic conductivity represents the ability of water to pass through the soil profile. K_x, K_y, and K_z represent the hydraulic conductivity in the X, Y, and Z directions, where K_x and K_y are the same in isotropic media. K_z is considered as 10% of the K_x and K_y (Sajeena., 2015). Specifically, for K_x in different geological layers, the ranges are as follows: Sandstone has a hydraulic conductivity ranging from 5×10^-13 to 10^-10, clay shows a range of 9×10^-7 to 6×10^-3, sand is 1×10^-11 to 4.7×10^-9, clayey sandstone ranges from 2.5×10^-13 to 5×10^-11 and granite gneisses range from 1×10^-13 to 2×10^-9.

Boundaries Conditions

The initial head values were employed to establish the initial distribution of heads for the simulation. Various boundary conditions, such as constant head, rivers, canals, recharge, and evapotranspiration, were applied to the domain model. Figure 3 illustrates the representation of these diverse boundary conditions assigned within the domain model. The grids designated as constant head in the model signify a consistent hydraulic head that remains unchanged over time. The Bay of Bengal is the southern boundary of the study area. It was assigned as a constant head for this study and the location of the Bay of Bengal as shown in figure 3. In this study, the characteristics of the soil and the flow of rivers play a pivotal role in governing the flow of groundwater. The conductance of a river signifies the volume of water that infiltrates the aquifer. Specifically, the Godavari River and the Upputeru River discharge into the Bay of Bengal.

Figure 3. Boundary conditions, including Bay of Bengal and Godavari River, are assigned in the model.

To accurately assign river conditions, several key inputs are required, including the river stage, riverbed bottom elevation, and the conductance of the river (Sajeena., 2015). The water surface elevation of the river is denoted by the river stage, while the elevation of the riverbed material is designated as the riverbed bottom. The resistance encountered by the flow of water from the river into the groundwater is quantified as conductance. In this study, conductance values are determined using equation 4 (Sajeena., 2015).

(3)

where

C = conductance in m²d^-1

K = represents hydraulic conductivity in md^-1

L = length of the river in the cell (m)

W = width of the river in the cell (m)

M = thickness of the river.

Drains play an important role in mitigating waterlogging conditions by facilitating the lowering of the water table. It's important to note that if the groundwater level is already below the level of the drain, the drain will have no impact. To accurately assign drains, two crucial parameters are necessary i.e., drain elevation and conductance. The drain elevation represents the level at which the drain is set, while conductance quantifies the flow rate from the surrounding area to the drain. A significant portion of groundwater replenishment occurs through recharge processes. This replenishment arises from various sources, including rainfall, excessive irrigation, and artificial recharge techniques. This parameter, measured in millimeters per year (mm/year), can be estimated as 10% of the total annual rainfall or computed using a specific equation (Sajeena., 2015). For this study, the latter approach was adopted, and the equation used for the estimation is shown in equation 5.

(4)

where R and R_av represent areal recharge and average annual rainfall in cm respectively.

Model Calibration and Validation

The developed model undergoes a calibration process under steady-state conditions, then in a transient state using the groundwater data collected from all the well locations (8 wells are located in the low-land, 37 in the mid-land, and 11 in the high-land) from 2003 to 2007. This involves adjusting the hydraulic conductivity by the trial-and-error method to minimize the head difference between the predicted head and observed head at steady-state conditions. As the hydraulic conductivity was iteratively modified, the observed and predicted values were monitored. Eventually, the error value stabilizes at a certain point, signaling that the model has been effectively calibrated. This calibration process ensures that the model accurately reflects real-world groundwater conditions and enhances its reliability for subsequent simulations and predictions. The calibration model was further validated by adding all the groundwater levels from 2008 to 2017. The performance of the model is assessed through a validation process that involves comparing observed data with modeled data across all sites. This validation step ensures that the model accurately predicts groundwater dynamics. Mean Standard Error (MSE), Root Mean Square Error (RMSE), and Normalized Root Mean Square Error (NRMSE) were used to assess the calibrated and validated model’s performance.

(5)

(6)

(7)

where

h_o = observed head values

h_c = estimated head values

n = total number of observed data.

Model Prediction

Subsequently, after model validation, the study advanced to forecast groundwater levels in the study area along with the lift irrigation project. This assessment was conducted through simulations utilizing the validated model, comparing scenarios with and without the implementation of the lift irrigation project. These simulations provide valuable insights into the potential changes in groundwater dynamics resulting from the irrigation project, aiding in informed decision-making and sustainable water resource management.

Results and Discussions

Spatial-Temporal Variation of Precipitation

The spatial distribution of long-term average annual precipitation over the WGD is presented in figure 4. The descriptive statistics for mean monthly and annual precipitation data are presented in table 2. The study area is categorized into three sections based on elevation: low-land (elevation up to 7.5 m), mid-land (elevation between 7.5 and 75 m), and high-land (elevation exceeding 75 m). Among the 44 weather stations used in this study, 14 are situated in low-land, 25 in mid-land, and 5 in high-land regions. Annual precipitation showed a strong spatial pattern across the study area, with precipitation gradually decreasing from the northeast part of the study area to the southwest. There was an approximate difference of 262 mm in average annual precipitation between northeast and southwest. The long-term annual average precipitation over the study area ranged from 877 mm to 1250 mm, with a mean of 1064 mm (table 2). In general, in the mid-land, precipitation is usually a limiting factor for crop production. The long-term average annual precipitation for the mid-land is 897 mm (SD = 1838 mm). The WGD usually receives most of its precipitation in summer (June to October). The mean monthly precipitation peaks in August (221 mm) and gradually decreases toward winters, with a long-term minimum precipitation of 3 mm observed in the month of January. The standard deviation (SD) of monthly precipitation increased gradually from January toward the summer months, peaked at 320 mm in October, and decreased again gradually toward November and December. The coefficient of variation (CV) showed an opposite trend to SD and was highest during winter months (80% in January) and lowest in the summer. On the other hand, a higher CV was observed for the mid-land, indicating a higher degree of spatial pattern in precipitation.

Figure 4. Variation of average annual rainfall across the study area from 2004 to 2017.

To further study the precipitation variation, monthly precipitation was plotted in a time series, along with a 12-month moving average, as shown in figure 5. The variation of monthly precipitation across the study area was shown in figure 5a, whereas the variation of average annual precipitation among low-, mid-, and high-land were shown in figures 5b, 5c, and 5d. The highest average annual precipitation of 1659 was recorded in 2010, which is one of the wettest years in the WGD. The lowest amount of annual precipitation was observed in 2009 as 646 mm, which was 39% lower than the long-term average annual precipitation. This trend of maximum average annual precipitation was consistent across the low-, mid-, and high-land. The highest average annual precipitation was recorded as 1916 mm in the high-land, whereas the lowest annual precipitation of 629 mm was recorded for the low-land. Overall, a decreasing trend was observed in the annual precipitation over the study area, with annual precipitation decreasing at a rate of 74 mm from 2004 to 2017. Decreasing precipitation trends were also observed for different regions, with rates of 134 mm, 132 mm, and 65 mm in the low-, mid-, and high-land, respectively, from 2004 to 2017. This declining trend could have a significant effect on the rise in pressure for existing water resources, which subsequently affects agriculture (Nair and Indu, 2020). A similar decreasing trend in precipitation was observed by Hengade and Eldho (2019) because of climate change. They used the grid-wise detrended meteorological data from 1981 to 2010 and observed an average change of -0.93% in annual precipitation in the Godavari River basin. Likewise, a decreasing trend in annual precipitation in the Godavari Middle Sub-basin was observed by Masroor and Sajjad (2024). They used the climate data from the Indian Meteorological Department from 1981 to 2017 and performed the non-parametric Mann-Kendall test and observed a decreasing trend in monthly precipitation.

Spatio-Temporal Variation of Groundwater Table

The spatial and temporal variation of groundwater studies was conducted using the measured groundwater data from observation wells and piezometers. The dynamics of groundwater levels are influenced by two primary factors: groundwater recharge (during rainfall, ephemeral river flow) and subsurface flow, in addition to the properties of the aquifer and soil (Wu et al., 1996; Lu et al., 2014). Areas with potential groundwater resources were identified considering the groundwater variation over time and space along with slope, land use and land cover, pumping details, and groundwater fluctuations map during the study period.

Temporal Variation

Temporal variation of groundwater resources in the study area during the study period was investigated by analyzing the change in groundwater levels during the study period. Average groundwater levels recorded in the observation wells located across the study area are shown in figure 6. Standard deviations for the 14-year period are also included for each data point to present the temporal variability for each observation well. A few wells located in the study area recorded major variation in the high-land whereas the observation wells located in the low- and mid-land recorded moderate fluctuation. It was observed that the highest variation in groundwater levels from 2003 to 2017 was recorded in Koyyalagudem (SD = 4.3 m), located in the high-land, whereas the lowest was recorded in Narsapuram (SD = 0.43 m). The higher SD reflects variation in climate conditions (precipitation) and management practices (irrigation extraction) between the years. Groundwater levels were less than a meter from the surface in two locations, i.e., in Narsapuram and Bhimavaram. To further investigate, a location was selected from each region—low-land, mid-land, and high-land—namely Akiveedu, Mogalthuru, and Buttayagudem, respectively, and the groundwater levels in these regions were plotted, as shown in figure 7. This clearly shows the variation of groundwater resources in different regions located across the study area.

Figure 5. Temporal variation of average annual precipitation in the (a) study area and (b) low-land, (c) mid-land, and (d) high-land within the West Godavari district.

Seasonal analysis is one of the best ways to investigate the variation of groundwater resources. Temporal variation of groundwater resources was majorly influenced by various parameters, including agricultural practices, overexploitation of groundwater resources, climate change, and soil properties. Thus, the difference between pre-monsoon and post-monsoon during the study period along with the precipitation data were plotted in figure 7 for low-, mid-, and high-land regions to study the influence of precipitation over the groundwater levels (Ghosh and Bera, 2023). The difference between groundwater levels between pre- and post-monsoon was recorded lowest in 2016, 2014, and 2011 in low-, mid-, and high-land regions, respectively. During this period, the rise of groundwater levels from pre-monsoon to post-monsoon in the low-, mid-, and high-land regions was recorded as 0.07, 0.48, and 0.67 m, respectively, which is very minimal. This could be due to below-average precipitation recorded in 2016 (744 mm) in the low-land, in 2014 (870 mm) in the mid-land region, and in 2011 (862 mm) in the high-land. Groundwater levels experienced the highest fluctuation in 2012, across low-, mid-, and high-land, where precipitation was recorded as 1609 mm, 1522 mm, and 1451 mm, respectively. This results in a rise in groundwater levels up to 1.33, 3.11, and 4.87 m in low-, mid-, and high-land regions, respectively. Following, there was little raise in 2014. Although precipitation has a major impact on groundwater levels, the degree of influence can vary depending on factors such as land management practices, topography, climate, soil type, and the soil's infiltration capacity (Smerdon, 2017). The long-term effects of pumping on groundwater resources varied by region, being minimal in the low-land region and higher in mid- and high-land regions. This could be due to the influence of surface water bodies in the low-land compared to mid- and high-land (Machiwal et al., 2012). The mid- and high-land regions are characterized by steeper slopes and higher drainage density, which limit infiltration and increase runoff during heavy rainfall events. This can lead to more pronounced subsurface water fluctuations compared to low-land regions. Additionally, land use changes and vegetation dynamics during this period could have further influenced the observed variations. This pattern aligns with findings by Mohan et al. (2018), which highlight the significant role of surface water bodies in sustaining groundwater recharge in low-land regions compared to mid- and high-land regions. Hussain et al., (2022) have correlated the groundwater response to precipitation while estimating the aquifer properties. This study estimated that annual recharge rates of 244–1472 mm year^-1 (12%–43%) of precipitation. Overall, the groundwater levels were decreasing in the study area. Similar decreasing trend was observed in the lower Godavari basin was observed by Haseena et al. (2015).

Figure 6. Average groundwater levels in the observation wells located in the study area.

Spatial Variation

Spatial variation of groundwater resources is mainly due to geological processes and other parameters, including aquifer properties, soil properties, topography, land use and land cover, water bodies interaction, and groundwater exploitation (Singh et al., 2010; Kulabako et al., 2007). These are the major parameters that could significantly affect the groundwater resources. The spatial variability of the groundwater system across the study area was investigated by comparing pre-monsoon and post-monsoon season groundwater maps for the years 2003, 2005, 2007, 2009, 2011, 2013, 2015, and 2017 developed using the kriging technique (Sharma and Irmak, 2012; Machiwal et al., 2012). The analysis focused on understanding the variation in groundwater levels between these two seasons, pre-monsoon in figure 8 and post-monsoon in figure 9.

Figure 7. Temporal variation of groundwater level difference (red line) between pre-monsoon and post-monsoon seasons in (a) low-land (b) mid-land and (c) high-land regions and its relation to annual precipitation (green bars).

The study further categorized groundwater level variations into five zones, with ranges of 0–2 m, 2–4 m, 4–6 m, 6–8 m, and 8–14 m for both pre- and post-monsoon seasons. On average major variation in the groundwater table was observed over the study area. Observations from the groundwater level maps revealed that in the pre-monsoon season, the high-regions like Dwaraka Tirumala exhibited lower groundwater levels compared to other areas. The depth of the water table was shallower along the seacoast and near rivers. Post-monsoon season depth-to-groundwater level maps indicated a rise in the groundwater table, resulting in a shift in contour positions. The average pre-monsoon level in the study area was 3.94 m, which changed to 2.21 m in the post-monsoon season. However, the rate of groundwater fluctuation varied across the low-land (0.83 m), mid-land (1.74 m), and high-land (2.62 m) regions. This variation within the study area could be due to the uneven distribution of rainfall, terrain, slope, and soil type. Additionally, surface water bodies such as Kolleru Lake, the Godavari River, and canals play an essential role in recharging the coastal and low-land regions. The seasonal streams and drainage networks provide significant water infiltration opportunities, particularly in flat terrains, which contribute to shallower water tables (Ramaiah et al., 2017; Appels et al., 2016; Monsieurs et al., 2015). On the other hand, mid- and high-land regions with limited surface water availability compared to low-land regions and with steep slopes have experienced reduced recharge potential. The average annual precipitation received in 2009 was 673.6 mm and 927.8 in 2016, where the water level rise from pre-monsoon and post-monsoon levels was less in low-land compared to mid- and high-land regions. Similar results were observed by Haseena et al. (2015).

Figure 8. Seasonal dynamics of groundwater during pre-monsoon variations in 2003, 2005, 2007, 2009, 2011, 2013, 2015, and 2017, respectively.

Figure 9. Seasonal dynamics of groundwater during post-monsoon variations in 2003, 2005, 2007, 2009, 2011, 2013, 2015, and 2017, respectively.

As per the Groundwater Scenario Report" released by the Andhra Pradesh government, in the previous thirty years, the number of wells in Andhra Pradesh and Telangana states alone has surged from 0.8 million to 2.5 million, and the area under groundwater irrigation has grown from 1 million to 3.4 million hectares (Sarkar, 2022; Kumar and Vedantam, 2016). According to a Central Ground Water Commission assessment, the basin's groundwater level is decreasing as a result of excessive extraction and insufficient or irregular precipitation (Sarkar, 2022; Patel et al., 2020). The impacts of climate change, characterized by susceptibility, vulnerability, and inefficient water resource utilization, are anticipated to escalate the reliance on groundwater, resulting in a decline in groundwater levels.

Figure 10 illustrates the variations in average groundwater levels across low-, mid-, and high-land regions during both pre-monsoon and post-monsoon periods. In the low-land region, the average groundwater level during the pre-monsoon period was around 2.5 meters below the surface, which then rose to 1.5 meters deep in the post-monsoon period. In contrast, the mid- and high-land regions recorded pre-monsoon groundwater levels at 4.5 meters and 4.9 meters deep, respectively, which decreased to 2.8 meters and 2.2 meters deep in the post-monsoon period for both regions.

Groundwater Fluctuation Map

The difference between successive rises and falls in the water level in observation wells over a year is referred to as fluctuation, where the rise can be attributed to recharge and the fall is a result of discharge. Using ArcGIS Pro, a groundwater table fluctuation map was developed for the study area. The average groundwater fluctuation for the period 2003 to 2017 was calculated and is shown in figure 11.

Figure 10. Variation of groundwater levels in pre- and post-monsoon.

Figure 11. Groundwater fluctuation map between 2003 and 2017.

Low fluctuation is denoted by yellow, medium fluctuation by orange, and high fluctuation by red. Analyzing the groundwater fluctuation map reveals that the hilly regions of the study area experience the maximum fluctuation in groundwater levels (4–5), while the low altitudes and areas closer to the sea exhibit minimum fluctuation (0–1 m). This spatial contrast in fluctuation highlights the diverse hydrogeological characteristics within the study area, highlighting the influence of topography and proximity to water bodies on groundwater dynamics. Haseena et al. (2015) reported findings aligning with the present study, noting that increased fluctuations in groundwater levels in hilly regions could be due to limited recharge potential and steep slopes.

Groundwater potential zones (GWPZs) were identified considering the slope, drainage density, pumping details, spatial and temporal variation, and fluctuation map for lithology data (Gaur et al., 2011). This comprehensive approach ensures that the model reflects the diverse factors influencing groundwater potential and recharge zones. The lithological data were obtained from the Department of Groundwater and Water Audit, Government of Andhra Pradesh. By analyzing the lithology of the study area, it is evident that sandstone and clay constitute the predominant top layer in most parts. Sandstone occupies a significant portion of WGD. Various locations in the study area exhibit a mix of clay, sand, clayey-sandstone, and granite-gneisses. The bottom layer typically consists of clayey sandstone and granite gneisses. Detailed information on lithological, soil, drainage, flow fluctuation, and DEM maps is provided in the Figshare link located in the supplementary material section later (fig. S1). Analysis revealed the division of the study area into four GWPZs, such as low, medium, moderately high, and high (fig. 12). The medium GWPZ covered the largest area at 3829 km², followed by the moderately high/high GWPZ with 3510 km², and the low GWPZ covered an area of 403 km². Additionally, the study revealed that around 740 villages were located within the medium GWPZ, followed by 436 villages in the moderately high to high GWPZ and 47 villages in the low GWPZ. Most of the moderately high to high GWPZ regions are located close to the water bodies, including the Bay of Bengal, Godavari River, Kolleru lake (freshwater lake), and Errakalva drain. The low GWPZ were further used to construct the Pattiseema lift irrigation project in 2017 by connecting the Godavari and Krishna rivers through an unlined canal to further enhance the recharge capacity of low GWPZs (fig. 12). Kamaraju et al. (1996) conducted a study to identify the potential zones and categorized zones into 11 classes. Some of the zones are located close to Kolleru Lake. This reflects the influence of surface water bodies to make groundwater resources sustainable. According to Prakash et al. (2014), Andhra Pradesh state, which is located in the Godavari River basin, was divided into four zones depending on the proportion of groundwater exploitation: safe (<70%), semi-critical (70% to 90%), critical (90% to 100%), and overexploited (>100%). Rajesh et al. (2021) also presented a study in which they identified the potential zones using weighted overlay analysis, the analytical hierarchy process (AHP), geographical information systems (GIS), and remote sensing. The results they obtained were consistent with the current study.

Model calibration

In this study, the groundwater model was calibrated under both steady-state and transient conditions, with a particular emphasis on the influential properties of hydraulic conductivity, recharge rates, and boundary conditions for ensuring model accuracy. This calibration process was achieved through a trial-and-error method.

Steady State Calibration

The crucial step in the calibration process is achieving a strong agreement between observed and measured hydraulic heads under steady-state conditions. In this step, the model was calibrated considering not much change in the hydraulic head (Guiger and Franz, 1996). Hydraulic conductivities of each layer were obtained from the groundwater department and were used in the beginning. Fine-tuning the hydraulic conductivity of the different layers in the soil profile was crucial in enhancing accuracy, which was accomplished through a trial-and-error approach (Mayer et al., 2007; Honnannanavar et al., 2023). The outcomes of this steady-state calibration are shown in figure 13, demonstrating the reasonable match between the observed and simulated initial water levels with an MSE of 2.17 m, an RMSE of 16.41 m, and an NRMSE of 12.06% was obtained between observed and predicted heads. Similar MODFLOW performance was observed by Tahershamsi et al. (2018) in the Aedebil Plains in Northwest Iran. They used quantitative models in terms of the main parameter affecting surface water changes using MODLOW and observed 2% relative normal root mean square error. They also indicated that the accuracy of the MODFLOW model was higher than the geostatistical methods due to the distribution of uncertainty in a few available piezometric points in the geostatistical method. Hani et al. (2007) developed a groundwater model aimed at assessing the risk assessment due to industrial effluent discharges. The model was calibrated in steady state and considered root mean square error (3.6 m) as a mount to determine the model accuracy.

Figure 12. Groundwater potential zones across the study area.

Transient State Calibration

The transient calibration was conducted for the period up to 2012, utilizing 11 discrete time steps in the simulation. During this process, adjustments were made to the storage and aquifer properties to minimize the RMSE. After numerous trial runs, involving variations in input and output parameters, the computed water levels demonstrated a reasonably close match to the observed values. The model was calibrated in a transient state with an R² of 0.84, as shown in figure 14. The MSE, RMSE, and NRMSE for the calibrated transient model are 1.93 m, 16.21 m, and 12.89%, respectively. Prasad and Rao (2018) changed the aquifer parameters to obtain the best NRMSE (11) for the model, which is close to the current study. They identified that the increase in groundwater dependency in the southern region has been increasing stress on existing groundwater resources and recommended the consideration of artificial recharge techniques as a viable solution to mitigate this stress.

Model Validation

Validation of the calibrated model is imperative to assess its predictive capability and instill greater confidence in its accuracy. In this study, groundwater levels up to 2017 were employed for the validation process. Incorporating the 'Pattiseema Lift Irrigation Project' into the model for validation is crucial. A recharge layer was introduced to account for this canal-induced recharge. To investigate the impact of the lift irrigation project, the model was also run without it, allowing for a comparative analysis of variations. Following validation, a comparison between observed and calculated heads for the model reveals a commendable agreement, as illustrated in figure 15. The MSE, RMSE, and NRMSE for the validated model are 1.1 m, 8.14 m, and 6.47%, respectively. Hernández et al. (2012) conducted a study using MODFLOW to predict groundwater levels in the irrigated fields in the Calera Aquifer region of central Mexico. They assessed model accuracy by varying model parameters and observed RMSE values of 25.1 m and 25.9 m between observed and predicted heads. The study emphasized that while the model demonstrated reasonable predictive capacity, further refinement and calibration of parameters would enhance its reliability for future groundwater predictions.

Figure 13. Calculated water levels vs. observed water levels for calibration under steady-state condition.

Figure 14. Calculated water levels vs. observed water levels for calibration under transient conditions.

Model Prediction

The validated model was employed to simulate groundwater levels to study the groundwater levels in the study area and understand the variation of groundwater levels. Modeled groundwater levels at each location in low-, mid-, and high-land regions were plotted as shown in figure 16. Projections were made for observation wells located in low-land (OB#1: Kaikaram), mid (OB#2: Pedatadepalli), and high-land (OB#3: Jangareddigudem) and are shown in figure 16. In 2018, the anticipated groundwater levels displayed distinctive patterns. OB#1 experienced an initial drop, followed by a period of stability. Groundwater levels in OB#2 had a slight increase after three years. On the other hand, OB#3 experienced a continuous fluctuation in groundwater levels. The location of groundwater wells was selected to ensure an accurate representation of the natural hydrogeological conditions in the region, i.e., low-, mid-, and high-land regions (Prakash et al., 2014). Pagnozzi et al. (2020) employed the MODFLOW model to predict groundwater flow in an unconfined aquifer, yielding comparable findings to those observed in the current study.

To further examine the impact of the 'Pattiseema Lift Irrigation Project' in the West Godavari, the model was run without the lift irrigation project. The recharge area was selected with the approximation in the model, and model-simulated groundwater levels were plotted for the observation well located in one of the locations close to the unlined canal, as shown in figure 17. Notably, the year 2017 saw a noticeable rise in groundwater levels due to the implementation of the lift irrigation project, particularly evident along the course of the Pattiseema canal through Pedatadepalli. Haseena et al. (2015) observed that the groundwater recharge rate is observed more in the lower basin than in the upper river basin, while overexploitation could lead to critical issues with water availability and result in seawater intrusion.

Figure 15. Observed vs. predicted heads plotted for the model with Pattiseema lift irrigation project.

Figure 16. Model predicted water levels in Kaikaram (low-land region), Pedatadepalli (mid-land region), and Jangareddigudem (high-land region).

This model was further used to predict the groundwater level heads for the next fifteen years with an increase of 5% pumping rate in the lift irrigation project for every five years with water available throughout the period and a decrease in recharge of 5%, and model predictions at an observation well in low-, mid-, and high-land are shown in figure 18. Water level fluctuations were recorded higher in the high-land region throughout the period when compared to low-land and mid-land. On the other hand, water levels in OB# 2 recorded a rise in water levels, whereas OB#3, which is located close to the surface water bodies, maintained close to the surface. Groundwater levels increased in areas with water availability in the canal and nearby water bodies. This can vary with the change in need for water resources, surface water availability, and with the overexploitation of groundwater resources. Overall, as a direct result of the lift irrigation project, groundwater levels in the canal area experienced a notable increase of around 3 meters in the OB#2 closely situated to the unlined canal. Another study conducted in WGD, Andhra Pradesh, using SWAT MODFLOW, shows that groundwater level increases due to canal irrigation (Venkata et al. 2022; Loukika et al., 2020; Surinaidu et al., 2014). From this study, WGD remains safe due to the river interlinkage project, and it was also found that Visual MODFLOW can be used efficiently for the evaluation and prediction of groundwater resources. Similarly, Chakraborty et al. (2020) used the validated model to predict the groundwater levels with good accuracy. They also concluded that the water movement from the Bay of Bengal to the study area indicates seawater intrusion.

Quantifying water resources and assessing water quality are very essential for optimal water usage and environmental sustainability. These insights are instrumental for future development and strategic planning to prevent potential groundwater crises in the region. However, the current limitations in the number of observation wells and data availability are the major constraints to studying the water resources in the study area. Therefore, it is critical to increase the number of groundwater monitoring wells and enhance water quality assessments, especially in coastal regions where the dependency on groundwater resources is growing rapidly. This data can further be used to assess the saltwater intrusion risks and rate. Likewise, the data and information obtained for this research will further be used to understand the effect of future climate scenarios on regional groundwater resources. Additionally, as wells for irrigation increased three times in the past 30 years, there is an urgent need to monitor the contamination of water resources and surface water bodies from fertilizers and pesticides applied for farming, which has critical implications for human health.

Figure 17. Predicted heads for Pedatadepalli with and without Pattiseema lift irrigation project.

Figure 18. Future predicted groundwater levels in observation wells located near Kaikaram (low-land region), Pedatadepalli (mid-land region), and Jangareddigudem (high-land region).

Conclusion

This study was conducted to comprehensively investigate the groundwater system in the WGD, with a specific focus on assessing the impact of the Pattiseema lift irrigation project in the study area. The average annual precipitation from 2004 to 2017 is 1064 mm. The study area was divided into three regions, i.e., low-, mid-, and high-land, and the average annual rainfall was recorded as 1113 in low-land, 897 in mid-land, and 989 mm in the high-land region. The identification of GWPZs involved a spatial and temporal analysis, resulting in the development of groundwater zone maps that captured both spatial and temporal variations. The study categorized the region into low, medium, moderately high, and high land GWPZs. The moderately high to high GWPZ are located close to the water bodies, including the Bay of Bengal, Godavari River, Kolleru Lake (freshwater lake), and Errakalva drain. The low GWPZ regions were further used to construct the Pattiseema lift irrigation project in 2017 by connecting the Godavari and Krishna rivers through an unlined canal to further enhance the recharge capacity of low GWPZs.

A numerical model was developed by inputting and calibrating the data, initially in steady-state conditions, then, in transient conditions, achieving an R² of 0.85 (MSE = 2.17 m, RMSE = 16.41 m, and NRMSE = 12.06%) and 0.84 (MSE = 1.93, RMSE = 16.21 m, and NRMSE = 12.89%), respectively. The validated model, with an R² of 0.90, MSE of 1.1 m, RMSE of 8.14 m, and NRMSE of 6.47%, was then used to predict groundwater levels. The effect of the Pattiseema lift irrigation project was integrated into the model to study the effect of the project on the rise of water resources. The simulated model revealed the effect of the unlined canal in the study area, showcasing a rise of approximately 3 meters. This observation signifies that the unlined canal has a positive impact, enhancing groundwater resources in proximity. This outcome emphasizes the critical importance of well-planned water management initiatives, emphasizing their role in sustaining and improving local hydrogeological conditions. The results indicate that the WGD benefits from the river interlinkage project, contributing to the overall security of groundwater resources in the area. The utilization of Visual MODFLOW has proven effective in evaluating and forecasting groundwater availability, enabling more informed decision-making for future resource management. The study's comprehensive approach serves as a valuable model for similar research endeavors, offering detailed insights into groundwater system conditions.

Supplemental Material

The supplemental materials mentioned in this article are available for download from the ASABE Figshare repository at: https://doi.org/10.13031/28392077

Acknowledgment

Authors appreciate the Groundwater and Water Audit Department, Government of Andhra Pradesh, India, for providing the data.

References

Adimalla, N., & Qian, H. (2021). Groundwater chemistry, distribution and potential health risk appraisal of nitrate enriched groundwater: A case study from the semi-urban region of South India. Ecotoxicol. Environ. Saf., 207, 111277. https://doi.org/10.1016/j.ecoenv.2020.111277

Akhter, S., & Hossain, M. S. (2017). Groundwater modelling of Dhaka City and surrounding areas and evaluation of the effect of artificial recharge to aquifers. World J. Res. Rev., 5(3), 54-60.

Appels, W. M., Bogaart, P. W., & van der Zee, S. E. (2016). Surface runoff in flat terrain: How field topography and runoff generating processes control hydrological connectivity. J. Hydrol., 534, 493-504. https://doi.org/10.1016/j.jhydrol.2016.01.021

Awais, M., Aslam, B., Maqsoom, A., Khalil, U., Ullah, F., Azam, S., & Imran, M. (2021). Assessing nitrate contamination risks in groundwater: A machine learning approach. Appl. Sci., 11(21), 10034. https://doi.org/10.3390/app112110034

Baghel, S., Tripathi, M. P., Khalkho, D., Al-Ansari, N., Kumar, A., & Elbeltagi, A. (2023). Delineation of suitable sites for groundwater recharge based on groundwater potential with RS, GIS, and AHP approach for Mand catchment of Mahanadi Basin. Sci. Rep., 13(1), 9860. https://doi.org/10.1038/s41598-023-36897-5

Chakraborty, S., Maity, P. K., & Das, S. (2020). Investigation, simulation, identification and prediction of groundwater levels in coastal areas of Purba Midnapur, India, using MODFLOW. Environ. Dev. Sustain., 22(4), 3805-3837. https://doi.org/10.1007/s10668-019-00344-1

Chiang, W. H. (2005). 3D-Groundwater modeling with PMWIN: A simulation system for modeling groundwater flow and transport processes. Heidelberg: Springer. https://doi.org/10.1007/3-540-27592-4

Chowdhury, A., & Rahnuma, M. (2023). Groundwater contaminant transport modeling using MODFLOW and MT3DMS: A case study in Rajshahi City. Water Pract. Technol., 18(5), 1255-1272. https://doi.org/10.2166/wpt.2023.076

Diersch, H. J. (2005). WASY software FEFLOW, finite element subsurface flow and transport simulation system, user’s manual. Berlin, Germany: WASY GmbH.

Directorate of Economics and Statistics. (2015). Handbook on statistics West Godavari District Andhra Pradesh. Government of India.

Gaur, S., Chahar, B. R., & Graillot, D. (2011). Combined use of groundwater modeling and potential zone analysis for management of groundwater. Int. J. Appl. Earth Obs. Geoinf., 13(1), 127-139. https://doi.org/10.1016/j.jag.2010.09.001

Ghosh, A., & Bera, B. (2023). Hydrogeochemical assessment of groundwater quality for drinking and irrigation applying groundwater quality index (GWQI) and irrigation water quality index (IWQI). Groundw. Sustain. Dev., 22, 100958. https://doi.org/10.1016/j.gsd.2023.100958

Guiger, N., & Franz, T. (1996). Visual MODFLOW: Users guide. Ontario, Canada: Waterloo Hydrogeologic, and Waterloo.

Hani, A., Djorfi, S., Djabri, L., Lamouroux, C., & Lallahem, S. (2007). Impact of the industrial rejections on water of Annaba aquifer (Algeria). Eur. Water, 19(20), 3-14.

Haseena, S., Surinaidu, L., & Giridhar, M. V. (2022). Assessment of groundwater and surface water resources in the Godavari Basin. Proc. 2nd National Conf. on Water Environment and Society.

Hengade, N., & Eldho, T. I. (2019). Relative impact of recent climate and land cover changes in the Godavari river basin, India. J. Earth Syst. Sci., 128(4), 94. https://doi.org/10.1007/s12040-019-1135-4

Hernández, J. E., Gowda, P. H., Howell, T. A., Steiner, J. L., Mojarro, F., Núñez, E. P., & Avila, J. R. (2012). Modeling groundwater levels on the Calera Aquifer Region in Central Mexico using ModFlow. J. Agric. Sci. Technol. B, 2(1), 52-61.

Honnannanavar, A., Patil, N., & Patil, V. (2023). Groundwater flow modeling of a microwatershed using visual MODFLOW flex. Curr. World Environ., 18(2). http://doi.org/10.12944/CWE.18.2.24

Hora, T., Srinivasan, V., & Basu, N. B. (2019). The groundwater recovery paradox in South India. Geophys. Res. Lett., 46(16), 9602-9611. https://doi.org/10.1029/2019GL083525

Hussain, F., Wu, R.-S., & Shih, D.-S. (2022). Water table response to rainfall and groundwater simulation using physics-based numerical model: WASH123D. J. Hydrol.: Reg. Stud., 39, 100988. https://doi.org/10.1016/j.ejrh.2022.100988

Kamaraju, M. V., Bhattacharya, A., Reddy, G. S., Rao, G. C., Murthy, G. S., & Rao, T. C. (1996). Ground-water potential evaluation of West Godavari District, Andhra Pradesh State, India - A GIS approach. Groundwater, 34(2), 318-325. https://doi.org/10.1111/j.1745-6584.1996.tb01891.x

Keery, J., Binley, A., Crook, N., & Smith, J. W. (2007). Temporal and spatial variability of groundwater-surface water fluxes: Development and application of an analytical method using temperature time series. J. Hydrol., 336(1), 1-16. https://doi.org/10.1016/j.jhydrol.2006.12.003

Kulabako, N. R., Nalubega, M., & Thunvik, R. (2007). Study of the impact of land use and hydrogeological settings on the shallow groundwater quality in a peri-urban area of Kampala, Uganda. Sci. Total Environ., 381(1), 180-199. https://doi.org/10.1016/j.scitotenv.2007.03.035

Kumar, C. P., & Seethapathi, P. V. (2002). Assessment of natural groundwater recharge in Upper Ganga Canal command area. J. Appl. Hydrol., 15(4), 13-20.

Kumar, G. P., & Kumar, P. A. (2014). Development of groundwater flow model using visual MODFLOW. Int. J. Adv. Res., 2(6), 649-656.

Kumar, M. D., & Vedantam, N. (2016). Chapter 7 - Groundwater use and decline in tank irrigation? Analysis from erstwhile Andhra Pradesh. In M. D. Kumar, A. J. James, & Y. Kabir (Eds.), Rural water systems for multiple uses and livelihood security (pp. 145-182). Elsevier. https://doi.org/10.1016/B978-0-12-804132-1.00007-X

Loukika, K. N., Venkata Reddy, K., Durga Rao, K. H., & Singh, A. (2018). Estimation of groundwater recharge rate using SWAT MODFLOW model. In J. K. Ghosh, & I. da Silva (Ed.), Applications of Geomatics in Civil Engineering: Select Proceedings of ICGCE 2018 (pp. 143-154). Springer. https://doi.org/10.1007/978-981-13-7067-0_10

Lu, W. X., Zhao, Y., Chu, H. B., & Yang, L. L. (2014). The analysis of groundwater levels influenced by dual factors in western Jilin Province by using time series analysis method. Appl. Water Sci., 4(3), 251-260. https://doi.org/10.1007/s13201-013-0111-4

Machiwal, D., Mishra, A., Jha, M. K., Sharma, A., & Sisodia, S. S. (2012). Modeling short-term spatial and temporal variability of groundwater level using geostatistics and GIS. Nat. Resour. Res., 21(1), 117-136. https://doi.org/10.1007/s11053-011-9167-8

Masroor, M., & Sajjad, H. (2024). Understanding climate change dynamics in the Godavari middle sub-basin using parametric and non-parametric models. Geosyst. Geoenviron., 3(2), 100269. https://doi.org/10.1016/j.geogeo.2024.100269

Mayer, A., May, W., Lukkarila, C., & Diehl, J. (2007). Estimation of fault-zone conductance by calibration of a regional groundwater flow model: Desert Hot Springs, California. Hydrol. J., 15(6), 1093-1106. https://doi.org/10.1007/s10040-007-0158-0

McDonald, M. G., & Harbaugh, A. W. (1988). A modular three-dimensional finite-difference groundwater flow model. US Geological survey techniques of water-resources investigation. Book 6, Chapter A1. US Geological Survey, Open-File report 83-875.

Meggiorin, M., Naranjo-Fernández, N., Passadore, G., Sottani, A., Botter, G., & Rinaldo, A. (2024). Data-driven statistical optimization of a groundwater monitoring network. J. Hydrol., 631, 130667. https://doi.org/10.1016/j.jhydrol.2024.130667

Merino, G. G., Jones, D., Stooksbury, D. E., & Hubbard, K. G. (2001). Determination of semivariogram models to krige hourly and daily solar irradiance in western Nebraska. J. Appl. Meteorol., 40(6), 1085-1094. https://doi.org/10.1175/1520-0450(2001)040<1085:DOSMTK>2.0.CO;2

Mirudhula, K. (2014). Impact of lined/unlined canal on groundwater recharge in the lower Bhavani Basin. Int. J. Eng. Res. Tech., 3(9), 1327-1329.

Mohan, C., Western, A. W., Wei, Y., & Saft, M. (2018). Predicting groundwater recharge for varying land cover and climate conditions – a global meta-study. Hydrol. Earth Syst. Sci., 22(5), 2689-2703. https://doi.org/10.5194/hess-22-2689-2018

Mondal, N. C., Singh, V. P., Singh, V. S., & Saxena, V. K. (2010). Determining the interaction between groundwater and saline water through groundwater major ions chemistry. J. Hydrol., 388(1), 100-111. https://doi.org/10.1016/j.jhydrol.2010.04.032

Monsieurs, E., Dessie, M., Adgo, E., Poesen, J., Deckers, J., Verhoest, N., & Nyssen, J. (2015). Seasonal surface drainage of sloping farmland: A review of its hydrogeomorphic impacts. Land Degrad. Dev., 26(1), 35-44. https://doi.org/10.1002/ldr.2286

Nair, A. S., & Indu, J. (2020). Changing groundwater storage trend of India after severe drought. Int. J. Remote Sens., 41(19), 7565-7584. https://doi.org/10.1080/01431161.2020.1763503

Nandimandalam, J. R., Sharma, K., & Alagappan, R. (2023). Preliminary investigation of saline water intrusion (SWI) and submarine groundwater discharge (SGD) along the south-eastern coast of Andhra Pradesh, India, using groundwater dynamics, sea surface temperature and field water quality anomalies. Environ. Sci. Pollut. Res., 30(10), 26338-26356. https://doi.org/10.1007/s11356-022-23973-y

Pagnozzi, M., Coletta, G., Leone, G., Catani, V., Esposito, L., & Fiorillo, F. (2020). A steady-state model to simulate groundwater flow in unconfined aquifer. Appl. Sci., 10(8), 2708. https://doi.org/10.3390/app10082708

Panneerselvam, B., Muniraj, K., Duraisamy, K., Pande, C., Karuppannan, S., & Thomas, M. (2023). An integrated approach to explore the suitability of nitrate-contaminated groundwater for drinking purposes in a semiarid region of India. Environ. Geochem. Health, 45(3), 647-663. https://doi.org/10.1007/s10653-022-01237-5

Patel, A., Rai, S. P., Akpataku, K. V., Puthiyottil, N., Singh, A. K., Pant, N.,... Noble, J. (2023). Hydrogeochemical characterization of groundwater in the shallow aquifer system of Middle Ganga Basin, India. Groundw. Sustain. Dev., 21, 100934. https://doi.org/10.1016/j.gsd.2023.100934

Patel, P. M., Saha, D., & Shah, T. (2020). Sustainability of groundwater through community-driven distributed recharge: An analysis of arguments for water scarce regions of semi-arid India. J. Hydrol.: Reg. Stud., 29, 100680. https://doi.org/10.1016/j.ejrh.2020.100680

Prakash, P., Kumar Das, A., & Sandilya, C. V. (2014). Study relating to the spatial distribution of ground water quality for different elements using legacy data of RWS&S, Andhra Pradesh, India. Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-8, 383–390. Retrieved from https://doi.org/10.5194/isprsarchives-XL-8-383-2014

Prasad, Y. S., & Rao, Y. R. (2018). Groundwater flow modelling of a micro-watershed in the upland area of East Godavari district, Andhra Pradesh, India. Model. Earth Syst. Environ., 4(3), 1007-1019. https://doi.org/10.1007/s40808-018-0502-5

Rajesh, J., Pande, C. B., Kadam, S. A., Gorantiwar, S. D., & Shinde, M. G. (2021). Exploration of groundwater potential zones using analytical hierarchical process (AHP) approach in the Godavari river basin of Maharashtra in India. Appl. Water Sci., 11(12), 182. https://doi.org/10.1007/s13201-021-01518-x

Ramaiah, S. N., Reddy, R., & Reddy, V. J. (2017). Over-exploitation of groundwater, and its impact on phreatic aquifer system, surface water bodies and rural water supply schemes in Malur Taluk, Kolar district, Karnataka - A case study. Nat. Environ. Pollut. Tech., 16(2).

Sajeena, S. (2015). Ground water resources modelling of a watershed using MODFLOW. PhD diss. Department of Land and Water Resources and Conservation Engineering, Kelappaji College of Agricultural Engineering and Technology.

Sarkar, S. (2022). Drought and flood dynamics of Godavari basin, India: A geospatial perspective. Arabian J. Geosci., 15(8), 772. https://doi.org/10.1007/s12517-022-10041-5

Sarkar, T., Kannaujiya, S., Taloor, A. K., Champati Ray, P. K., & Chauhan, P. (2020). Integrated study of GRACE data derived interannual groundwater storage variability over water stressed Indian regions. Groundw. Sustain. Dev., 10, 100376. https://doi.org/10.1016/j.gsd.2020.100376

Shah, T. (2020). Climate change and groundwater: India’s opportunities for mitigation and adaptation. In Water resources policies in South Asia (pp. 213-243). India: Routledge. https://doi.org/10.4324/9780367818487-14

Shankar, P. V., Kulkarni, H., & Krishnan, S. (2011). India’s groundwater challenge and the way forward. Econ. Polit. Wkly., 46(2), 37-45. Retrieved from https://www.jstor.org/stable/27918012

Sharma, V., & Bhambota, S. (2022). Strategies to improve crop-water productivity. In C. Ray, S. Muddu, & S. Sharma (Eds.), Food, energy, and water nexus: A consideration for the 21st century (pp. 149-172). Cham: Springer. https://doi.org/10.1007/978-3-030-85728-8_8

Sharma, V., & Irmak, S. (2012). Mapping spatially interpolated precipitation, reference evapotranspiration, actual crop evapotranspiration, and net irrigation requirements in Nebraska: Part I. Precipitation and reference evapotranspiration. Trans. ASABE, 55(3), 907-921. https://doi.org/10.13031/2013.41523

Sharma, V., Rudnick, D. R., & Irmak, S. (2013). Development and evaluation of ordinary least squares regression models for predicting irrigated and rainfed maize and soybean yields. Trans. ASABE, 56(4), 1361-1378. https://doi.org/10.13031/trans.56.9973

Siade, A. J., Hall, J., & Karelse, R. N. (2017). A practical, robust methodology for acquiring new observation data using computationally expensive groundwater models. Water Resour. Res., 53(11), 9860-9882. https://doi.org/10.1002/2017WR020814

Singh, A. (2022). Groundwater recharge assessment and long-term simulation for managing the threat of salinization of irrigated lands. J. Hydrol., 609, 127775. https://doi.org/10.1016/j.jhydrol.2022.127775

Singh, S., Singh, C., & Mukherjee, S. (2010). Impact of land-use and land-cover change on groundwater quality in the Lower Shiwalik hills: A remote sensing and GIS based approach. Open Geosci., 2(2), 124-131. https://doi.org/10.2478/v10085-010-0003-x

Smerdon, B. D. (2017). A synopsis of climate change effects on groundwater recharge. J. Hydrol., 555, 125-128. https://doi.org/10.1016/j.jhydrol.2017.09.047

Suhag, R. (2016). Overview of ground water in India. PRS On Standing Committee On Water Resources, Legislative Research.

Surinaidu, L., Gurunadha Rao, V. V., Srinivasa Rao, N., & Srinu, S. (2014). Hydrogeological and groundwater modeling studies to estimate the groundwater inflows into the coal Mines at different mine development stages using MODFLOW, Andhra Pradesh, India. Water Resour. Ind., 7-8, 49-65. https://doi.org/10.1016/j.wri.2014.10.002

Swathy, P. S. (2016). Simulation of saltwater intrusion into the coastal aquifers of Kadalundi river basin in Malappuram district using visual mudflow. PhD diss. Academy of Climate Change Education and Research.

Tahershamsi, A., Feizi, A., & Molaei, S. (2018). Modeling groundwater surface by MODFLOW math code and geostatistical method. Civil Eng. J., 4(4), 812. https://doi.org/10.28991/CEJ-0309135

Venkata Rao, G., Shiva Kumar, P., & Srinivas, K. (2022). Analysis of groundwater flow for saltwater interaction range using visual modflow: A case study of coastal belt in Srikakulam District of Andhra Pradesh, India. Asian J. Water Environ. Pollut., 19(4), 91-97. https://doi.org/10.3233/ajw220060

Venkatramanan, S., Viswanathan, P. M., & Chung, S. Y. (2019). GIS and geostatistical techniques for groundwater science. Elsevier.

Wingle, W. L., Poeter, E. P., & McKenna, S. A. (1999). UNCERT: Geostatistics, uncertainty analysis and visualization software applied to groundwater flow and contaminant transport modeling. Comput. Geosci., 25(4), 365-376. https://doi.org/10.1016/S0098-3004(98)00140-X

Wu, J., Zhang, R., & Yang, J. (1996). Analysis of rainfall-recharge relationships. J. Hydrol., 177(1), 143-160. https://doi.org/10.1016/0022-1694(95)02935-4