Keywords

1 Introduction

Groundwater is an integral part of the hydrological cycle. Groundwater is a key resource for human development. On a global scale, one-third of the population depends on groundwater for their drinking water, in urban as well as rural areas. Groundwater also plays a pivotal role in agriculture, and an increasing portion of groundwater is being used for agriculture. Central Ground Water Board mentioned that contribution of groundwater for irrigation was 48.19% in their 2009 report (Kambale et al. 2007). The IPCC (2013) has reported that changes in climatic factors will affect the groundwater availability and quality. Presently in most of the places in India groundwater is being over-exploited. This over-exploitation of water resources may result in future famine. Hence, it is necessary to maintain the groundwater availability for the future period.

The climate parameters temperature, precipitation, wind speed, and solar radiation have major effects on the groundwater recharge capacity. The change in trend pattern of climate variables gives the idea of the change of climate over the region. Many previous assessments had been focused on trend analysis of climate variables and change in groundwater level with climate change. Kumar [14] reviewed coupling of climate data with the specific hydrological model used for the study of water balance components. The spatiotemporal change of climate variables has its impact on the water balance component of the watershed and reported that the highest rainfall accelerated the recharge over the watershed [17]. Marhaento et al. [16] studied the change of hydrological components in response to future landuse change and climate change. With their study, it was concluded that the fraction of changes is more under the climate change scenario, so the impact of land-use change is slower compared to climate change. Mogaji et al. [18] assessed the recharge rate and PET with the empirical relationship equation. The change of climatic conditions concerning economic and environmental development scenarios such as A1B, A2, B1. Alam et al. [1] assessed the groundwater storage of California central valley using distributed variable infiltration capacity model (VIC), where the change in groundwater was assessed for the different cropping patterns and agriculture practice.

Shrestha et al. [30] assessed the change of groundwater level for the period of 2011–2100 using the WETPASS and MODLFOW model. They used average precipitation and temperature from five different GCM under two different RCP scenarios of 4.5 and 8.5. In their study, the recharge rate gets decreased mainly during the dry season. Kahsay et al. [12] assessed future groundwater recharge and base flow of Tekeze basin, Ethiopia using the WETSPA model. They concluded that the rate of precipitation has impact on variation of recharge rate. Kambale et al. [13] predicted the climate change impact on the northern part of India for the future period of 2030. The groundwater level variation was studied using the hydrological models HYDRUS-1D and MODFLOW. They reported that the recharge rate might vary with the number of climate parameters considered for analysis and soil types over the region. Shrestha et al. [30] assessed the climatic factor change for the period of 2006–2100 and its influence on recharge for four Asian cities. The climate parameters from five different RCM were coupled with the model WETSPASS and SWAT model. Based on the size of the area the model was selected and concluded that recharge is mainly due to amount of rainfall and temperature. More decrease in recharge occurs during the summer season due to an increase in temperature. Loukika et al. [15] estimated the groundwater recharge of chintalapudi village, West Godavari for the period of 1990–2000. For the groundwater recharge estimation coupled SWAT-MODFLOW were used, where they calculated the groundwater recharge on monthly basis in wells based on aquifer parameters.

In this paper, groundwater recharge variation for Nagavali basin is studied under climate change scenarios over 60 years from 2021 to 2080. The monthly groundwater recharge has been estimated using the SWAT model. The climate models namely ACCESS, CNRM, and MPI have been used for the climate data retrieved, and the groundwater recharge has been estimated using a model. The seasonal and annual variation of climate variables and groundwater recharge has been studied over the entire study basin.

2 Study Area

The study area was the Nagavali river basin shown in Fig. 1 is an imperative basin among the east-flowing river. An independent and interstate river lies in between the Mahanadi and Godavari basins. The Nagavali originates near the Lakhabahal in the Kalahandi district of Odisha at an elevation of about 1300 m (source: http://www.dowrodisha.gov.in/BasinMaps/IndexofBasins.htm). It drains parts of the districts of Odisha and Andhra Pradesh State. Basin lies on the geographical coordinates of 18° 17′ to 19° 44′ latitude and 82° 53′ to 83° 54′ longitude. The total basin area is 9510 Sq. Km. This river plays a major role in supporting irrigation demand and water supply demand for the district of Koraput, Rayagada, Vizianagaram, and Srikakulam. The total population of the basin as per census 2011 was 6, 76, 678. The top region of basin was covered with forest so that the lower portion of 5559Sq. Km was analyzed in the study.

Fig. 1
figure 1

Location map of Nagavali river basin

Full size image

3 Data and Methods

3.1 Data

For calculation of recharge of aquifer using SWAT model, the SWAT model requires weather parameters (precipitation and temperature), streamflow data, the digital elevation model, landuse and landcover, and soil properties.

The climate parameters (rainfall and temperature) for the historic period were downloaded from the Indian Meteorological Department (IMD), Pune. The rainfall data is available with the resolution of 0.25° × 0.25°. For the analysis, eight grid points data within the basin were downloaded for the period of 1990–2018. The temperature of 1°×1° grid was also collected for the period of 1990–2018. For the future rainfall and temperature, three climate models were selected data has been downloaded and bias-corrected for the future period of 2021–2080.

An Aster DEM of 30 m × 30 m shown in Fig. 2 was used to delineate watershed and stream network generation. The landuse/landcover (LULC) data and soil map of the study area was clipped from SWAT Indian datasets. LULC shown in Fig. 3 is useful for the generation of hydrological response units (HRUs). The soil map shown in Fig. 4 shows that the entire area has three types of soil namely loam, sandy loam, and sandy clay loam. The majority of the upstream portion is covered with sandy clay loam soil. The stream flow data obtained from the Srikakulam gauge station is essential for the calibration of the SWAT model.

Fig. 2
figure 2

Digital elevation model

Full size image
Fig. 3
figure 3

Landuse/landcover map

Full size image
Fig. 4
figure 4

Soil map

Full size image

3.2 Methodology

The methodology for the work includes the coupling of climate data with the hydrological model SWAT that is stated in Fig. 5. The study involves analysis of the spatiotemporal trend pattern of precipitation, temperature, and groundwater level. The first step is to analyze the trend of measured and downloaded climatic variables. Then using the geospatial and weather data SWAT model was initially running for the historic period of 1991–2018. In the SWAT model, the basin was subdivided into sub-basins for analysis. The model was calibrated and validated with the observed streamflow data of the Srikakulam gauge station using SWAT CUP. The climate model data from three models namely ACCESS, MPI, CNRM were used for future precipitation and temperature. The data has been downloaded under 2 RCP Scenarios RCP 4.5 and RCP 8.5 and the downloaded model data were bias-corrected using Quantile mapping. The bias-corrected climate variable was given as input for calibrated SWAT model. Using the SWAT HRU output file the deep percolation value, evapotranspiration amount for each sub-basin and each land-use type were calculated. Finally, the percentage change in evapotranspiration and groundwater recharge was calculated over 60 years for the prediction of future possibilities of recharge.

Fig. 5
figure 5

Conceptual flowchart for assessing the groundwater recharge

Full size image

4 Result and Discussion

4.1 Bias Correction

The climate data has been corrected for systematic error in the data using Quantile mapping method. The bias corrections of monthly precipitation for the base period of 1950–2005 are presented in Fig. 6. From the figure, corrected precipitation is found to be consistent with observed precipitation. Hence, the corrected data shows close agreement with the observed value in the historic period. So with the best fit transformation function, the future period (2021- 2080) rainfall has been corrected.

Fig. 6
figure 6

Correlation between observed and simulated rainfall

Full size image

4.2 Trend Analysis

4.2.1 Precipitation

MK test statistic (S) indicates a significant increasing trend in future precipitation. Rainfall is predicted to increase gradually in both the scenarios. From Fig. 7 the rainfall variation compared to base period shows that there is 6% increase in RCP 4.5 scenario and 85% increase in RCP 8.5 scenario. The average annual precipitation was 2192.93 and 1186.54 mm. There is a decrease in precipitation for the years 2024, 2025, and 2026 compared to the present average precipitation. Hence, in the future, these years may receive less rainfall.

Fig. 7
figure 7

Rainfall variation

Full size image

4.2.2 Temperature

For the base period, the maximum average temperature was 32.26 °C and the minimum average temperature was 20.92 °C. The average temperature increases by 2% and 4% under RCP 4.5 and RCP 8.5 scenarios. The temperature plot shown in Fig. 8 depicts that the temperature is more under RCP 8.5 scenario.

Fig. 8
figure 8

Trend plot of future temperature

Full size image

4.2.3 Evapotranspiration

The average evapotranspiration is 119.39 and 126.91 mm under RCP 4.5 and RCP 8.5 scenarios, respectively. The increase in temperature by 4% under RCP 8.5 is causing more evapotranspiration compared to RCP 4.5 scenario (Fig. 9).

Fig. 9
figure 9

Potential evapotranspiration

Full size image

4.3 Estimation of Groundwater Recharge

4.3.1 Calibration and Validation of Model

The SWAT model was calibrated with the observed monthly streamflow data with the simulated streamflow data using SWAT CUP. The model was calibrated for the period of 1991–2005 and validated for the period for 2006–2014. The statistical model performance indicators R2 and NSE for the period of the calibration were 0.85 and 0.80, respectively and for the period of validation was 0.73 and 0.70, respectively.

4.3.2 Water Balance Component and Groundwater Recharge

After running the calibrated SWAT model, the water balance components and groundwater recharge have been simulated for the future period of 2021–2080. The rainfall, evapotranspiration, and recharge variation have been presented in the table show that there is an increase in future precipitation with the increase of temperature by 0.46 and 1.08 °C under climate change scenarios (RCP 4.5 and RCP 8.5). However, there is a possibility of decrease in the rate of evaporation due to an increase in rainfall amount (Table 1).

Table 1 Percentage change of water balance components
Full size table

Therefore, the decrease in PET by 16% to 11% under RCP 4.5 and RCP 8.5 leads to an increase of recharge rate by 57–90%. The average groundwater recharge (Amount of water entering the deep aquifer) is 9.62 m of water and 11.64 m of water under scenarios 4.5 and 8.5. The reason for the increase in recharge is due to the permeability of the soil and the increase in precipitation rate. The combination plot of rainfall and recharge is shown in Fig. 10a, b indicates that the rainfall has direct influence on the recharge. Therefore, the year with higher rainfall has higher recharge capacity.

Fig. 10
figure 10

Trend plot of recharge and rainfall under a RCP 4.5, b RCP 8.5

Full size image

The spatial distribution map of recharge and evapotranspiration for both the scenarios was prepared in GIS environment under both the scenarios. The recharge map is shown in Fig. 11. It is seen that the future period (2021–2080) has possibility of higher recharge capacity over the basin under both scenarios. The evapotranspiration plot is shown in Fig. 12.

Fig. 11
figure 11

Spatial distribution of recharge in the basin

Full size image
Fig. 12
figure 12

Spatial distribution of evapotranspiration in the basin

Full size image

4.3.3 Temporal Analysis of Water Balance

The temporal analysis refers to the seasonal analysis of weather and water balance components. For seasonal analysis, the entire year was divided into three seasons which is presented in Table 2. Season I refers to the monsoon season (June to September), Season II refers to the winter season (October to January), and Season III is termed as the summer season (February to May).

Table 2 Season-wise water balance
Full size table

The season-wise analysis of the water balance is presented in Table 2. Under RCP 4.5 scenario, the monsoon season receives 76% of total rainfall, contributes 56% of total recharge, and 25% of evapotranspiration. However, in the summer season, the increase of temperature contributes to 40% of total evapotranspiration, 1% of total recharge. Hence, in the summer season, the recharge rate is limited with increased evapotranspiration with increased temperature.

Under RCP 8.5 Scenario, the monsoon season receives 70% of total rainfall, contributes 59% of total recharge with 25% of evapotranspiration. The rate of recharge is less and potential evapotranspiration is more during the summer season. Hence the increase in temperature is affecting the recharge condition. In the RCP 8.5 scenario, the percentage of winter rainfall is 27% that is more than RCP 4.5 winter rainfall and it indicates that there is, possibility of a shift in the rainfall if the likely emission scenario of the atmosphere is 8.5.

5 Conclusions

The study involves the change in recharge with the influence of climatic conditions. This study analyses the spatiotemporal patterns of weather, groundwater recharge, and evapotranspiration during the period of 2021–2080 under RCP scenarios of 4.5 and 8.5. Trend pattern of rainfall and temperature shows that it increases gradually in the upcoming years. There is a decrease in the future evapotranspiration by 16 and 11% under RCP 4.5 and 8.5 scenarios compared to the present status. From the simulation output, it is seen that the groundwater recharge is increasing in the future period. The average groundwater recharge for the study area in the base period was 6.13 m. In the future period, the average recharge over the basin was 9.62 and 11.64 m under RCP 4.5 and RCP 8.5, respectively and average evapotranspiration was 119.45 mm and 126.91 m. There is decrease in recharge during summer season with the increase of evapotranspiration rate. The increase in precipitation by 6–85% is causing the recharge to increase by 57–90% in the future period. Overall, in the Nagavali basin, there is increase in groundwater recharge in the upcoming years. The reason for the increase of recharge is increase in rainfall rate and soil with higher permeability.