India is one of the most drought‐ravaged countries in the world and faces at least one drought in one region or another in every 3 years. There is no single reliable approach in characterizing future droughts. To understand future drought risk, potential changes of drought properties and characteristics are analysed in this study. Using Fuzzy c‐means clustering approach, homogeneous drought regions are identified in the Godavari river basin and therefore, optimum number of clusters were assigned as four. The 12‐month standardized precipitation index (SPI) using precipitation data from India Meteorological Department (IMD) and Global Climate Model (GCM)—MIROC‐ESM‐CHEM is calculated for the homogeneous regions of the Godavari basin. The best fit copula for observed and simulated severity and duration are: Region 1—Clayton, Regions 2 and 3—Gumbel, Region 4—Frank copula. Severity‐duration‐frequency (SDF) and severity‐area‐frequency (SAF) curves were developed and analysed using the best fit copulas. The research findings conclude that moderate and severe droughts are frequently increasing for future periods (2006–2099) compared to the historic period (1962–2005). Droughts with high severity and high mean interarrival time are observed as expected in the future. For the Godavari basin, the SDF curves were concave upwards indicating an increase in severity with an increase in duration. The rate of increase of severity is small for shorter durations compared to that of longer‐duration drought. Thus, more prolonged drought events in the 21st century are likely to occur. The SAF curves with steeper slopes and high variability in topographical and hydrological characteristics have been observed over the Godavari basin. From these curves, for a specified percentage of area and return period, the drought severity can be calculated and the information can be used for crop management and agricultural water demands. Overall, the findings of this research offer a view of likely scenarios of drought in the Godavari basin.

In present study, MIKE 11NAM rainfall runoff(R-R) and MIKE BASIN models were used to develop R–R relationship and to analyze the water availability status of the Shipra river basin under changing hydrological conditions. The model was developed using discharge data observed for 9 years at the Ujjain G/d site. The coefficient of determination(R²) values for NAM R-R model calibration and validation were 0.685 and 0.642, respectively, indicating good agreement between the simulated and observed runoff. As the Shipra River is subjected to a number of hydrological interventions, the MIKE BASIN model was used to develop various scenarios to study the influence of river diversion, on-stream dam, river interlinking on water availability in different time-scale at desired locations. Scenario analysis of MIKEBASIN results indicated that the Khan diversion, which was executed to avoid entry of contaminated Khan river water into Shipra has reduced the water availability in the Shipra river. However, the simulation results indicated that the water availability in Shipra has been improved due to Narmada-Shipra link, a minimum of 1.7 m³/s flow is available at Ujjain even during the low-flow period. Gambhir reservoir was developed to evaluate how demand deficit can be fulfilled through proper reservoir operation during a drought period.

Because of changing climatic conditions, uneven distribution of rainfall occurs throughout India. As a result, dependence on groundwater for irrigation has increased tremendously for industrial and domestic purposes. In India approximately 89% of agricultural demands are met through groundwater. Due to increases in population, demand for groundwater and lack of effective utilization have resulted in rapid depletion of groundwater in most parts of the country. Therefore, quantifying groundwater resources is a serious concern in populated states of India, because it is now difficult to supply enough water to every citizen, and will remain so in the future. Because of difficulties in accessing observation data, researchers have begun to depend on satellite-based remote sensing information to deal with groundwater variations. The present study deals with filling the data gap between Gravity Recovery And Climate Experiment (GRACE) and GRACE Follow On (GRACE FO) missions using multilayer perceptron’s (MLPs) during 2017–2018 to obtain a continuous terrestrial water storage anomaly (TWSA) series from 2003 to 2020 for Telangana state, India. The MLP model performed well in predicting the TWSA, with a correlation coefficient of r = 0.96 between modeled TWSA and GRACE TWSA during the test period. Telangana state observed negative TWSAs (annual) in the years 2003, 2004, 2005, 2009, 2012, 2015, and 2016–19. This TWSA series (2003–2020) was then used to evaluate regional groundwater storage anomalies (GWSAs) in Telangana state, which is considered to be one of the water stress regions in India. The TWSAs were converted to GWSAs using Global Land Data Assimilation System (GLDAS) parameters. The Telangana state experienced decreasing GWSA in the years 2005, 2009, and 2012, and from 2015 to 2019, leading to severe droughts. Groundwater well measurements were obtained from the Central Groundwater Board (CGWB) and converted to GWSA at a seasonal scale. The GWSAs obtained from GRACE (GWSA_GRACE) were converted to seasonal values and compared with GWSAs obtained from observation well data (GWSA_obs). The performance metrics of r = 0.74, RMSE = 5.3, and NSE = $0.62$ were obtained between (GWSA_GRACE) and (GWSA_obs), representing a good correlation among them. Over the past decade, Telangana state has significantly relied on groundwater resources for irrigation, domestic, and industrial purposes. As a result, evaluating groundwater storage variations at a regional scale may help policy makers and water resource researchers in the sustainable utilization and management of groundwater resources.

Groundwater is a vital resource for domestic, agricultural, and industrial use, with its demand growing due to population growth and climate change. Several studies have identified groundwater depleting in India at unsustainable rate over North-west part, but a contrasting trend is observed in the southern India. To better study groundwater dynamics quality-controlled and reliable well data is essential, which is missing. Here we process seasonal groundwater levels from 32,299 wells across India to obtain reliable well data and provide respective specific yields. Initially, wells with no data and negative values are removed. Later three-sigma method is imposed on each well to eliminate outliers. Finally, wells with at least two values per year, with no value repeating more than twice consecutively, are retained, resulting in 2,759 reliable wells. We used vectorization-based method to classify aquifer types and estimated specific yields based on hydrogeological map. We also provide open access to data and scripts so that researchers can study groundwater variations, compare GRACE and model-based groundwater estimates against in-situ well data.

The Gravity Recovery and Climate Experiment (GRACE) satellite mission provides estimates of total water storage anomalies (TWSA) in terms of “mascons” (mass concentration blocks), representing integrated changes in surface water, soil moisture, and groundwater. However, the coarse spatial resolution of GRACE mascons (~3°) limits its utility for regional-scale hydrological analysis. Although several downscaling methods have been proposed to improve resolution, none have been comprehensively validated against in-situ observations. In this study, we are validating both native and downscaled GRACE TWSA products using well-based in-situ groundwater observations across India. Furthermore, we develop an improved downscaling method by modifying the approach of Vishwakarma et al. (2021), i.e., incorporating mass conservation constraints at the native GRACE resolution instead of catchment scale (Vishwakarma et al., 2021) to better capture spatial variability at a finer 0.5° grid scale. We assessed the efficacy of our downscaled product and two existing downscaled GRACE products (Vishwakarma et al., 2021; Gao and Soja, 2024) using performance metrics such as gain in correlation (r) and gain in root-mean-square error (RMSE) relative to native GRACE product. The entire India is covered by 50 mascons. Out of these 50 mascons groundwater observations are available only at 22 mascons. Our modified approach shows improved performance across 12 mascons in India, with median gains in r ranging from 0.58 % to 8.84 % and gain in RMSE improvements from 0.24 % to 23.59 %. Whereas, the Vishwakarma et al. (2021) and Gao and Soja (2024) methods perform well in only 3 and 10 mascons, respectively. These results demonstrate that our enhanced downscaling approach provides more accurate and spatially resolved estimates of groundwater storage changes, offering a valuable tool for regional water resource assessments in India. Further, the downscaled GRACE product developed in this study (Mascon wise Mass Conservation product) is publicly accessible at https://figshare.com/articles/dataset/GRACE_downscaled_TWSA_product_using_Mascon_wise_Mass_Conservation/29196617.