Article | Open

Groundwater rejuvenation in parts of India influenced by water-policy change implementation

  • Scientific Reports 7, Article number: 7453 (2017)
  • doi:10.1038/s41598-017-07058-2
  • Download Citation
Published online:


The dwindling groundwater resource of India, supporting almost one fifth of the global population and also the largest groundwater user, has been of great concern in recent years. However, in contrary to the well documented Indian groundwater depletion due to rapid and unmanaged groundwater withdrawal, here for the first time, we report regional-scale groundwater storage (GWS) replenishment through long-term (1996–2014, using more than 19000 observation locations) in situ and decadal (2003–2014) satellite-based groundwater storage measurements in western and southern parts of India. In parts of western and southern India, in situ GWS (GWSobs) has been decreasing at the rate of −5.81 ± 0.38 km3/year (in 1996–2001) and −0.92 ± 0.12 km3/year (in 1996–2002), and reversed to replenish at the rate of 2.04 ± 0.20 km3/year (in 2002–2014) and 0.76 ± 0.08 km3/year (in 2003–2014), respectively. Here, using statistical analyses and simulation results of groundwater management policy change effect on groundwater storage in western and southern India, we show that paradigm shift in Indian groundwater withdrawal and management policies for sustainable water utilization appear to have started replenishing the aquifers in western and southern parts of India.


Groundwater is the largest liquid freshwater resource of the Earth. It plays a crucial role in human sustenance and global food security by supporting irrigated agriculture1. At present, India (Fig. 1a) is undergoing a “groundwater drought”2. The country comprises <3% of the terrestrial area and hosts about 19% of the global population. It also covers more than 30% of the global irrigated land2 and consumes the largest volume of global groundwater resource (higher than the sum of the total groundwater abstraction of United States and China, the second and third countries, respectively, in the country-wise groundwater utilization list)3. The country is witnessing a rapid rise in population, urbanization and change in anthropogenic water use, cropping pattern and lifestyle leading to unsustainable abstraction of available groundwater4,5,6,7 (e.g. 245 billion cubic meters (BCM) irrigational groundwater abstracted from India alone during 2011 only)8, which is at least 25% of the total global groundwater withdrawal3. These result to groundwater withdrawal to availability ratio being higher than 0.8 (i.e. more than 80% of the available groundwater has been withdrawn) in most parts of the country9. The country has been placed in the top of the list of groundwater depletion (GWD) with 33.9% of the global GWD linked with food production and trade10. In recent summers several parts of the country have witnessed law-and-order situation linked with groundwater availability11. At least 54% of India has been identified to be highly to extremely water stressed12, with present water demand of 712 BCM, projected to be increasing to 833 BCM in 2025 and 899 BCM in 205013. Accordingly, the Indian groundwater scenario has become a global paradigm for future availability and resilience to human strategies. Future groundwater management is challenging with new socio-political alignments at present and near future, and impending, potential climate change2.

Figure 1
Figure 1

(a) The study area including groundwater level (GWL) trend (calculated based on culled, temporally continuous groundwater level measurements [n = 4316] between 1996 and 2014); (b) annual mean precipitation (mm/year) between 1979 and 2014. Rectangular outlines indicate the five hydro-meteorological zones (A to E) delineated based on the duration of the hydro-meteorological seasons (monsoon, post-monsoon and pre-monsoon); (c) map of different aquifer types, classified based on hydraulic conductivity (K). Major rivers are marked by blue lines; (d) map of significant correlation coefficient (r, p value 0.05) between GWSobs and GWSsat; Maps of trends of (e) positive (blue), and (f) negative (red) GWSsat anomalies, respectively. All the maps were made using Ferret program (NOAA), QGIS software47 and standard graphical illustrators.

Causes of groundwater depletion also include inefficient water use practices, wasteful irrigation systems with poor maintenance, and low prices of both power for electricity-driven well pumps and supplied water14. Further, the natural groundwater availability and recharge in the region is extremely heterogeneous because of diversity in hydrogeologic set-up15, 16 and climatic conditions15 (Fig. 1b,c). While, much of the northern India is characterized by the highly groundwater-enriched, porous aquifers of the Indus-Ganges-Brahmaputra (IGB) river basins15,16,17 that are considered as the “bread-basket” of South Asia, much of peninsular India is composed of low-yielding, crystalline aquifers15, 16. Similarly, even as the country’s average precipitation is relatively high (~120 cm/year)18, it is highly variable both spatially (Asian monsoonal path dependent) and temporally (>75% precipitation during the four monsoon months)18, resulting to the Indian climate varying from extremely arid to some of the wettest places on Earth (Fig. 1b). Thus, the entire study area has to be classified into five different Hydro-meteorological zones (HMZ) with similar climatic pattern (see methods and supplementary information).

However, contrary to the aforesaid condition of groundwater depletion, general observations suggest groundwater replenishment trends in some parts of India, with groundwater storage (GWS) renewal trends of up to ~4 m rise in groundwater level in recent years19, specifically in parts of the western Indian state of Gujarat (HMZ B), and south Indian state of Andhra Pradesh (HMZ E, undivided state [including present Telengana state] indicated in Fig. 1a). While, unregulated abstraction for enhanced irrigation of water-intensive cultivation (e.g. boro rice, including Basmati)19 is resulting to one of the most rapid and drastic groundwater depletion in human history3, recent paradigm shift in Indian central/state government management strategies on groundwater withdrawal and stress (e.g. Pradhan Mantri Krishi Sinchayee Yojana) will likely start to demonstrate its results in near future. Policies like restriction of subsidized electricity for irrigation20, separate electricity distribution for agricultural purpose (Jyotigram Yojana)21, construction of large-scale, regional enhanced recharge systems in water-stressed crystalline aquifers (e.g. ~700 million USD allocated to Tapti river mega recharge project22), artificial recharge of 85 BCM/year in ~0.5 million km2 through ~10 million structures23, enhanced recharge by interlinking of river catchments (e.g. Narmada-Sabarmati interlinking24), will probably start replenishing the aquifers25 by increasing groundwater storage in near future. To quantify the replenishment trends, we used two-decade (1996–2014) long in situ and decadal (2003–2014) satellite-based measurements of groundwater conditions. We also used robust statistical approaches and a global-scale hydrological model simulation results in order to show the influence of water management policy change (based on data availability) in replenishment of groundwater storage in western and southern parts of the country.

Result and Discussion

Groundwater storage estimates

We used seasonal in situ groundwater storage anomalies (GWSAobs) and monthly groundwater storage anomalies (GWSAsat, 2003–2014), calculated using the Gravity Recovery and Climate Experiment (GRACE) satellite mission and land surface model-simulated soil moisture and surface water equivalents, to quantify the groundwater storage anomaly trends in the Indian region (Fig. S2a–l). The two anomalies demonstrate good correlation with each other (p value < 0.05) (Fig. 1d), and were found to substantiate each other for the overlapped time period (detailed information on validation of satellite-based products in the study region can be found in Bhanja et al.16). Trend analyses of groundwater level anomalies (GWLA), GWSAobs (1996–2014) and GWSAsat (2003–2014) indicate increasing trends in western (HMZ B) and southern (HMZ E) India, respectively (Figs 1a,e and 2), corroborating well with the aforesaid general observations of groundwater rejuvenation. GWSAobs data indicate renewal of GWS in HMZs B and E at a rate of 1.06 ± 0.03, and 0.31 ± 0.02 km3/year, on the other hand, the HMZs A and D have been subjected to rapid GWS depletion at a rate of 4.55 ± 0.11 km3/year and 3.59 ± 0.14 km3/year, respectively (Fig. 2). Satellite-based observation shows, the western (HMZ B) and southern (HMZ E) India, are experiencing GWSAsat renewal at a rate of 0.53 ± 0.30 and 0.69 ± 0.27 cm/year (2.29 ± 1.32 and 3.90 ± 1.50 km3/year, with p value < 0.05), respectively (Fig. S2a–l). Concurrently, rapid groundwater depletion (Fig. 1f) was observed in northern (HMZ A) and eastern (HMZ E) India, respectively, at a rate of 1.40 ± 0.14 and 1.16 ± 0.35 cm/year (14.02 ± 1.37 and 14.49 ± 4.36 km3/year, p < 0.05). These observations are in agreement with previous in situ observations over IGB basin (depletion at a rate of ~8 km3/year between 2000 and 2012)17, and satellite-based observations of GWS depletion from the general areas, e.g. northwest India (17.7 km3/year, 2002–08)4 and northern (54 km3/year, 2002–08)5 India, Ganges basin (1.25 cm/year, 2003–14)7 and Bangladesh (located within HMZ D)6.

Figure 2
Figure 2

Maps of annual in situ groundwater storage anomalies (GWSAobs) over the study area between 1996 and 2014. GJ (Gujarat) and AP (Andhra Pradesh) indicate the location of specific study areas. All of the maps were made using QGIS software47.

Groundwater storage rejuvenation in specific study areas

In order to find the potential effects of groundwater management policies, long term, temporally continuous, groundwater level data (1996–2014) from specific study areas of western India (Indian state Gujarat, n = 177 in situ locations) and southern India (Indian state Andhra Pradesh, n = 350 in situ locations) were studied in details. The GWSAobs were decomposed between trends and cycles using non-parametric Hodrick-Prescott (HP)26 filter approach. The HP filter analyses show declining groundwater trends from (pre-)1996 time, with GWSAobs decreasing at the rate of 5.81 ± 0.38 km3/year (linear trend), and reversal to increasing trends (GWSAobs increasing at the rate of 2.04 ± 0.20 km3/year) from 2002 in Gujarat (Figs 3a and S5). Similarly, declining trend (0.92 ± 0.12 km3/year) of Andhra Pradesh from (pre-)1996 reversed around 2003, increasing at the rate of 0.76 ± 0.08 km3/year, respectively (Figs 3a and S6). The trend reversals are found to be synchronous with various groundwater policy changes in these two states. For the same time period (i.e. 1996–2014), there were no visible changes in HP trends of precipitation for the two study areas (Fig. 3a and b), thus reducing possibilities of much influence of rainfall on the aforesaid trends of groundwater storage increase or decrease.

Figure 3
Figure 3

Annual GWSAobs, GWSAsat, and their Hodrick-Prescott (HP) trends for (a) Gujarat (in HMZ B) and (b) Andhra Pradesh (in HMZ E). The grey area represents the time-period for implementation of groundwater management policy change. Change in simulated groundwater abstraction from PCR-GLOBWB simulation for Gujarat is shown in (a). Precipitation HP trends are shown using black line.

Influences of water management policy change on groundwater storage

Reduction in agricultural power supply from 16 billion units to 10 billion units between 2001 and 200627 might results in reduction in irrigational groundwater quantities (as electricity-driven wells are responsible for over 90% of the groundwater withdrawal)21. The groundwater withdrawal is assumed to be directly proportional to the agricultural electricity usage that is required to operate the pump (see methods). We tried to understand the potential effects of these policy changes (reduction of agricultural electricity usage, AEU) on groundwater storage anomaly in Gujarat between 1996 and 2012 (based on data availability) by using first-order Bayesian Vector Autoregression (BVAR). We have used Minnesota prior for our estimation, which automatically takes care of the potential non-stationarity of the data. Based on the BVAR estimate, we have also generated Impulse Response Functions (IRFs) to uncover the dynamics of AEU and GWSAobs. We have used Cholesky Decomposition for one standard deviation shock to innovation and ordering AEU to GWS to generate the IRFs (Fig. S4). Results show electricity usage for groundwater irrigation was found to have significant (p value < 0.05) negative impact on GWSAobs in Gujarat, implying that reduction of electricity usage leads to increase in GWSAobs. We have also performed the Granger causality analyses28,29,30, in order to study the causal relationship between reduction in electricity usage and GWSAobs. Result indicates AEU reduction Granger causes GWSAobs in Gujarat. In order to show the effects of policy change on GWS in Gujarat, we have also used a global-scale hydrological model, PCR-GLOBWB31, 32, based on available policy change information. The model was operated in a Baseline Scenario (BS) using observed meteorologic and other model forcing parameters (i.e. precipitation, irrigation data, domestic and industrial water use etc.; see methods and supplementary information). An Hypothetical Test Scenario (HTS) has been considered, which includes BS scenario along with updated groundwater policy change information. We used the electricity reduction information in HTS simulation of Gujarat (i.e. lowering ~1/3 of agricultural power consumption) and compared the output GWS with that of the BS. The simulation results support our observation with an estimated increase in GWS by 3.2 to 4.4 km3/year in Gujarat (p < 0.05) during 2003–2014 on comparing the GWS for 1996–2002. It is interesting to note that the total precipitation rate in the area has not increased, even decrease in mean annual precipitation has been observed (linearly decreasing at a rate of 9.11 ± 4.45 mm/year) during 2002–2014.

Similarly, in Andhra Pradesh, UN FAO programs between 2004 and 2008, trained farmers on groundwater sustainable management practises33. Prior to the UN FAO program, another project named APWELL34 was linked with groundwater development in Andhra Pradesh. Substantial increase in surface water irrigation (SWI) have been observed 1996 onwards (>18% increase in SWI between 2003–2011 from 2002 level)35, thereby possibly inducing enhanced non-meteoric recharge36. The IRFs has been studied using first-order BVAR using Minnesota prior. We have used Cholesky Decomposition for one standard deviation shock to innovation and ordering SWI to GWSAobs for 1 standard deviation innovations shock (Fig. S5) between 1996 and 2011 (based on data availability). Results show significant (p value < 0.01) positive impact of SWI on GWSAobs (Table S2, Fig. S5). It is also found that SWI Granger causes GWSAobs, thus supporting our hypothesis of groundwater rejuvenation in AP from 2004 onward, being possibly associated with enhanced SWI.

Assumptions and limitations

We inferred the increasing trends of groundwater to be related to groundwater management strategy adaptation in the detailed study areas (e.g. decreasing power subsidy or increasing artificial recharge by creating surface water bodies), through advanced statistical approaches and simulation of a global-scale hydrological model. In present knowledge scenario, impacts of some of the water management issues are almost impossible to quantify, for example, the effect of change in farmer management practice on groundwater resources in Andhra Pradesh etc. Total groundwater abstraction (Q) for Gujarat can be computed following Eq. 1 (see methods). It can be said that, groundwater abstraction is directly proportional to the electricity consumption. The information has been incorporated in PCR-GLOBWB to simulate change in groundwater storage in Gujarat. The policy related information are available from 1996 to 2012 (Gujarat) and 2011 (Andhra Pradesh), hence, the analyses are conducted up to the time period of data availability. The PCR-GLOBWB simulations are extended for two more years i.e. 2013 and 2014 using the values of 2012.


We conclude that in India, where huge groundwater consumption is widely known to be leading to severe dwindling of groundwater resource in recent times, previously unreported, discernable GWS replenishment can also be observed in certain Indian regions. Specifically, in parts of the western (HMZ B) and southern (HMZ E) India, GWSobs decreased at the rate of −5.81 ± 0.38 km3/year (in 1996–2001) and −0.92 ± 0.12 km3/year (in 1996–2002), and reversed to replenish at the rate of 2.04 ± 0.20 km3/year (in 2002–2014) and 0.76 ± 0.08 km3/year (in 2003–2014), respectively. This groundwater storage rejuvenation may possibly be attributed to implementation of ingenious groundwater management strategies in both Indian public and private sectors. We have tested and substantiated this hypothesis in the two areas in western and southern India, by Bayesian VAR estimates, causality test and simulating the effects of water management policy changes, as applicable, by using a global hydrological model. Thus, while the northern and eastern parts of India are still undergoing acute usable groundwater depletion and stress, encouraging, replenishing groundwater scenarios are detected in the western and southern India under proper water resource management practices.


Study area

India has extremely heterogeneous spatial climatic pattern with variable precipitation and other meteoric water component (e.g. humidity). To overcome this meteorological heterogeneity, for this study, the region was divided into 5 hydro-meteorological zones (HMZ) on the basis of their precipitation and specific humidity pattern. We used long term monthly mean precipitation data set [Global Historical Climatological Network (GHCN)] from the year 1960 to 2010 for >120 locations spread over the study area. We filtered on availability of at least 70% of continuous data. As a result, precipitation data over 37 locations were used in the present study. We used European Center for Medium Range Weather Forecasting (ECMWF) reanalysis37 (ERA Interim) simulation of specific humidity (SH) in the present study area between 1979 and 2012 to constrain the boundary of the hydro-meteorological seasons (Supplementary Information) corresponding to different hydro-meteorologic zones (HMZs) (Supplementary Information). We used monthly meteorological sub-division-wise in situ precipitation data for Gujarat and Andhra Pradesh between 1996 and 201438. We used 10 × 10 monthly gridded precipitation data from ECMWF’s simulation of reanalyzed (ERA Interim)37 total precipitation product from 2003 to 2014 to prepare the Fig. 1b.

In situ groundwater level measurements

We have collected groundwater level (GWL) data from Central Ground Water Board (CGWB, India) repository for measurements between January 1996 and November 2014 from mostly unconfined aquifers (~87%)39 for a total number of observation wells in 19278 locations, having GWL data collection of maximum four times a year (late post-monsoon [January], pre-monsoon [May], monsoon [August] and early post-monsoon [November]). In order to use the continuous data for anomaly analyses, we have selected the locations with at least three seasonal data in every year and outliers were removed following Tukey’s fence approach40, reduced the usable well numbers to 4316. Linear trend in water level data were calculated using a linear regression model17. GWL anomaly (GWLA) for each of these wells was computed after subtraction of all time mean depth to water table from absolute depth of water table in each season and the sign made reversed for depth below the surface convention. The GWLA information has been transformed to GWS anomalies (GWSAobs) after incorporating specific yield information for each well16, 18 (details can be found within Supplementary information). The median values of GWSAobs within a grid cell were used to make 10 × 10 gridded data for the whole study area and compared with GWSAsat.

Satellite-based groundwater storage estimation

We used 133 monthly GRACE RL05M mascon solutions, obtained from the Jet Propulsion Laboratory, National Aeronautics and Space Administration (NASA)16, 41 between January 2003 and December 2014, to determine terrestrial water storage (TWS) anomaly. Anomalies were calculated related to the mean during the entire period. In order to determine groundwater storage (GWS) anomalies, anomalies of other water components of terrestrial water cycle i.e. soil moisture (SM), and surface water (SW) equivalents were removed from TWS anomalies. We used the Global Land Data Assimilation System (GLDAS)42 for SM and SW estimates in the corresponding 133 GRACE measurement months over the study period. We used combination of three different models’ output, Community Land Model (CLM2), Variable Infiltration Capacity (VIC), and NOAH to remove any bias associated with a single model. Before comparing with GWLA and GWSA, we used GRACE based estimates corresponding to the observational time period only i.e. four times a year between 2005 and 2013. Subsequently, we removed 36-monthly mean from TWS, SM, and SR of a particular month and followed the same procedure discussed above for calculating GWSsat anomaly.

Quantification of influence of policy change on groundwater

The non-parametric trend analysis approach, the Hodrick-Prescott (HP) statistical filter26 is a robust technique used to decompose time series data into trend and cycle (see SI for more details). It allows data to reveal its own trend and does not impose any trend structure on the data arbitrarily. As a result it can even capture nonlinearity in trend if it is present in the data. HP filter was used in trend analysis of GWL in the high resolution study area (Gujarat and Andhra Pradesh). We have also used first order Bayesian vector auto-regression (VAR) analysis43 and test for Granger causality28 for GWL and GWI in GJ and SWI in AP, in order to show causal relationships. A best practice parameter (BP) was also introduced in AP to account for the farmers training on sustainable groundwater management practice.

Total groundwater abstraction (Q) for Gujarat can be computed following the equation36:

Q= N × r × t A

where, N = total number of pumps, r = pumpage rate, t = time and A = geographical area. Assuming N and r are fixed over time, reduction in electricity consumption would actually lead to decrease in t.

Error analysis

We have estimated one-sigma trend error in the GRACE TWS anomaly estimates (σTWS). Errors associated with soil moisture anomaly (σSM) and surface water anomalies (σSR) were determined by estimating the standard deviations of the trends of the three GLDAS models. Subsequently, error estimates in the trend of GWS anomaly (σGWS) are estimated following the equation,

σ GWS =[ ( σ TWS ) 2 + ( σ SM ) 2 + ( σ SR ) 2 ]

Errors in GWSobs and precipitation data were obtained by estimating one-sigma trend error for calculating the linear trend.

Hydrological model simulation

We have simulated the effects of policy change using a widely used, global-scale hydrological model, PCR-GLOBWB, that simulates surface water and groundwater resources at 0.10 × 0.10 global grids (i.e., roughly 10 km by 10 km at the equator) on a daily time step31, 32. The model simulates water fluxes (e.g., runoff, infiltration, percolation) and water storage (e.g., soil moisture, groundwater), and allows water exchange in two vertically stacked soil layer and an underlying groundwater layer, which is the part of deeper soil layers that is not directly linked with vegetation. The groundwater layer was directly fed by active recharge through upper soil layers. Groundwater recharge was balanced by capillary rise if the water table lies within 5 m below the ground surface31, 44. Aquifers were parameterized based on lithology and topography of the region and represented as linear reservoir model45. Observation based WATCH ERA-Interim (WFDEI) was used in PCR-GLOBWB for climate forcing46, the observed precipitation data has been used here. More information on model description can be found in supplementary information.

Additional Information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Aeschbach-Hertig, W. & Gleeson, T. Regional strategies for the accelerating global problem of groundwater depletion. Nature Geoscience 5, 853–861 (2012).

  2. 2.

    Goldin, T. Groundwater: India’s drought below ground. Nature Geoscience 9, 98–98 (2016).

  3. 3.

    Margat, J. & Van der Gun, J. Groundwater around the world: a geographic synopsis. CRC Press (2013).

  4. 4.

    Rodell, M., Velicogna, I. & Famiglietti, J. S. Satellite-based estimates of groundwater depletion in India. Nature 460, 999–1002, doi:10.1038/nature08238 (2009).

  5. 5.

    Tiwari, V. M., Wahr, J. & Swenson, S. Dwindling groundwater resources in northern India, from satellite gravity observations. Geophys. Res. Lett. 36, L18401 (2009).

  6. 6.

    Shamsudduha, M., Taylor, R. G. & Longuevergne, L. Monitoring groundwater storage changes in the highly seasonal humid tropics: Validation of GRACE measurements in the Bengal Basin. Water Resour. Res. 48, W02508 (2012).

  7. 7.

    Panda, D. K. & Wahr, J. Spatiotemporal evolution of water storage changes in India from the updated GRACE-derived gravity records, Water Resour. Res. 51 (2015).

  8. 8.

    Central Ground Water Board (CGWB), G.o.I., Ministry of Water Resources, 2014a. Dynamic Ground Water Resources of India.

  9. 9.

    Bates, B. C. et al. Climate Change and Water. Technical Paper of the Intergovernmental Panel on Climate Change Intergov. Panel on Clim. Change, Secr. Geneva, Switzerland (2008).

  10. 10.

    Dalin, C., Wada, Y., Kastner, T. & Puma, M. J. Groundwater depletion embedded in international food trade. Nature 543, 700–704 (2017).

  11. 11.

    The Hindu (2016). accessed on 07 July, 2016.

  12. 12.

    WRI, 2015 accessed on 20 September, 2016.

  13. 13.

    UNICEF, Water in India: situation and Prospects (2013).

  14. 14.

    Department of Agriculture and Cooperation, G.o.I., Ministry of Agriculture. State of Indian Agriculture 2012–13, p. 221pp (2013).

  15. 15.

    Mukherjee, A. et al. Groundwater systems of the Indian Sub-Continent. Journal of Hydrology: Regional Studies 4, 1–14 (2015).

  16. 16.

    Bhanja, S. N., Mukherjee, A., Saha, D., Velicogna, I. & Famiglietti, J. S. Validation of GRACE based groundwater storage anomaly using in-situ groundwater level measurements in India. Journal of Hydrology 543, 729–738 (2016).

  17. 17.

    MacDonald, A. M. et al. Groundwater quality and depletion in the Indo-Gangetic Basin mapped from in situ observations. Nature Geoscience 9, 762–766 (2016).

  18. 18.

    National Climate Centre, India Meteorological Department. Monsoon report 2012, (2013).

  19. 19.

    Sidhu, H. S. Production conditions in contemporary Punjab agriculture. JPS 12, 197–217 (2005).

  20. 20.

    Barker R. et al. Managing irrigation in an environment of water scarcity. Rice in the Global Economy: Strategic Research and Policy Issues for Food Security. International Rice Research Institute, Los Banos, the Philippines (2010).

  21. 21.

    Shah, T. & Verma, S. Co-management of electricity and groundwater: An assessment of Gujarat’s Jyotirgram scheme. Economic and Political Weekly 43, 59–66 (2008).

  22. 22.

    News article on Tapti river recharge project, accessed on Mar 21, 2016.

  23. 23.

    Central Ground Water Board (CGWB), G. o. I., Ministry of Water Resources. Master Plan for Artificial Recharge to Ground Water in India, p. 225 (2013).

  24. 24.

    Gujarat govt website accessed on 25 September, 2016.

  25. 25.

    News article on Artificial Recharge Scheme, accessed on Mar 21, 2016.

  26. 26.

    Hodrick, R. J. & Prescott, E. C. Postwar US business cycles: an empirical investigation. Journal of Money, credit, and Banking, 1–16 (1997).

  27. 27.

    Socio Economic Review 2012–13, Gujarat State, G. o. Gujarat, p. 233 (2013).

  28. 28.

    Granger, C. W. Some recent development in a concept of causality. Journal of econometrics 39, 199–211 (1988).

  29. 29.

    Tuttle, S. & Salvucci, G. Empirical evidence of contrasting soil moisture–precipitation feedbacks across the United States. Science 352, 825–828 (2016).

  30. 30.

    Russo, T. A. & Lall, U. Depletion and response of deep groundwater to climate-induced pumping variability. Nature Geoscience 10, 105–108 (2017).

  31. 31.

    Wada, Y., Wisser, D. & Bierkens, M. F. P. Global modeling of withdrawal, allocation and consumptive use of surface water and groundwater resources. Earth Syst. Dynam. 5, 15–40 (2014).

  32. 32.

    Wada, Y., de Graaf, I. E. & van Beek, L. P. High‐resolution modeling of human and climate impacts on global water resources. Journal of Advances in Modeling Earth Systems 8, 735–763 (2016).

  33. 33.

    Food and Agriculture Organization of the United Nations. Andhra Pradesh Farmer Managed Groundwater Systems, 55 pp. (2008).

  34. 34.

    Judicious management of groundwater through participatory hydrological monitoring – A manual by APWELL, 60pp. (2003).

  35. 35.

    Socio Economic Survey 2013–14, Andhra Pradesh, G. o. Andhra Pradesh, p. 367 (2014).

  36. 36.

    Mukherjee, A., Fryar, A. E. & Howell, P. D. Regional hydrostratigraphy and groundwater flow modeling in the arsenic-affected areas of the western Bengal basin, West Bengal, India. Hydrogeology Journal 15, 1397–1418 (2007).

  37. 37.

    Dee, D. P. et al. The ERA-Interim reanalysis: Configuration and performances of the data assimilation system. Quarterly Journal of the Royal Meteorological Society 137, 553–597 (2011).

  38. 38.

    India meteorological department (IMD). accessed on 10 September, 2016.

  39. 39.

    Central Ground Water Board (CGWB), G. o. I., Ministry of Water Resources. Groundwater year book 2013–14, p. 76 (2014).

  40. 40.

    Tukey, J. W. Exploratory Data Analysis. Addison-Wesley (1977).

  41. 41.

    Watkins, M. M. et al. Improved methods for observing Earth’s time variable mass distribution with GRACE using spherical cap mascons. J. Geophys. Res. Solid Earth 120 (2015).

  42. 42.

    Rodell, M. et al. The Global Land Data Assimilation System. Bull. Am. Meteorol. Soc. 85, 381–394 (2004).

  43. 43.

    Litterman, R. B. Forecasting with Bayesian vector autoregressions—five years of experience. Journal of Business & Economic Statistics 4, 25–38 (1986).

  44. 44.

    Bhanja, S. N., Rodell, M., Li, B., Saha, D. & Mukherjee, A. Spatio-temporal variability of groundwater storage in India. Journal of Hydrology 544, 428–437 (2016).

  45. 45.

    Kraijenhoff van de Leur, D. A. A study of non-steady ground-water flow with special reference to the reservoir-coefficient. De Ingenieur 19, 87–94 (1958).

  46. 46.

    Weedon, G. P. et al. The WFDEI meteorological forcing data set: WATCH Forcing Data methodology applied to ERA-Interim reanalysis data. Water Resour. Res. 50, 7505–7514 (2014).

  47. 47.

    QGIS Development Team. QGIS Geographic Information System. Open Source Geospatial Foundation Project (2013).

Download references


SNB acknowledges CSIR (India) for their support for providing SPM fellowship. SNB also acknowledges U.S. Department of State for the Fulbright fellowship. We acknowledge National Climatic Data Center (NCDC) and Central Ground Water Board (Ministry of Water Resources, River Development and Ganga Rejuvenation, Government of India) for availability of open-source GHCN and water level measurement data, respectively. The ERA Interim data were obtained from the ECMWF data distribution portal GRACE land data are available at, supported by the NASA MEaSUREs Program. The GLDAS data used in this study were acquired as part of the mission of NASA’s Earth Science Division and archived and distributed by the Goddard Earth Sciences (GES) Data and Information Services Center (DISC). The authors acknowledge the advice and insights provided by Dr. Dipankar Saha, CGWB for this study. SNB also thanks Shubha Verma (IIT Kharagpur, India) and Richard G. Taylor (UCL, UK) for their advice. The authors wish to acknowledge use of the Ferret program for analysis and graphics in this paper. Ferret is a product of NOAA’s Pacific Marine Environmental Laboratory (Information is available at The authors also acknowledge the professional editorial processing by the handling editor and constructive reviews of the two anonymous reviewers. They also QGIS software (, version 2.2.0, for some of the graphics used in this paper.

Author information


  1. Hydroscience and Policy Advisory Group, Department of Geology and Geophysics, Indian Institute of Technology, Kharagpur, WB, India

    • Soumendra N. Bhanja
    •  & Abhijit Mukherjee
  2. Hydrological Sciences Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD, USA

    • Soumendra N. Bhanja
    •  & Matthew Rodell
  3. School of Environmental Science and Engineering, Indian Institute of Technology, Kharagpur, West Bengal, 721302, India

    • Abhijit Mukherjee
  4. International Institute for Applied Systems Analysis, Laxenburg, Austria

    • Yoshihide Wada
  5. Department of Physical Geography, Utrecht University, Utrecht, The Netherlands

    • Yoshihide Wada
  6. Hydroscience and Policy Advisory Group, Department of Humanities and Social Sciences, Indian Institute of Technology, Kharagpur, WB, India

    • Siddhartha Chattopadhyay
  7. Department of Earth System Science, University of California, Irvine, USA

    • Isabella Velicogna
  8. Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA

    • Isabella Velicogna
    •  & James S. Famiglietti
  9. School of Physical Sciences, University of California, Irvine, USA

    • Kishore Pangaluru
    •  & James S. Famiglietti


  1. Search for Soumendra N. Bhanja in:

  2. Search for Abhijit Mukherjee in:

  3. Search for Matthew Rodell in:

  4. Search for Yoshihide Wada in:

  5. Search for Siddhartha Chattopadhyay in:

  6. Search for Isabella Velicogna in:

  7. Search for Kishore Pangaluru in:

  8. Search for James S. Famiglietti in:


S.N.B. and A.M. performed the background analyses and designed the study. S.N.B. accomplished the study under supervision of A.M. and advices from M.R. S.N.B. performed the data retrieval, management and processing. PCR-GLOBWB simulations were performed by Y.W. S.N.B. and S.C. have been involved in the statistical analyses. S.N.B. and A.M. wrote the manuscript with inputs from M.R., Y.W., S.C., I.V., K.P. and J.F.

Competing Interests

The authors declare that they have no competing interests.

Corresponding author

Correspondence to Abhijit Mukherjee.

Electronic supplementary material


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Creative Commons BY

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit