The availability of accurate rainfall data at proper temporal and spatial scales is vital for knowledge of renewable water resources and safe withdrawals for irrigation. Rain gauge networks in mountainous basins such as the Indus are sparse and insufficient to plan withdrawals and water management applications. Satellite rainfall estimates can be used as an alternative source of information but need area-specific calibration and validation due to the indirect nature of the radiation measurements. In this study, a calibration protocol is worked out for rainfall data from the Tropical Rainfall Measuring Mission (TRMM) satellite because uncalibrated TRMM rainfall data are inaccurate for use in rainfall–runoff studies and in soil water balance studies. Two alternative techniques, regression analysis (RA) and geographical differential analysis (GDA), were used to calibrate TRMM rainfall data for different periods and spatial distributions. The validity of these techniques was tested using Nash–Sutcliffe efficiency and the standard error of estimate. The GDA technique proved to be better, with higher efficiency and smaller error in complex mountainous terrains. The deviation between TRMM data and rain gauge data was decreased considerably from 10.9% (pre-calibration at 625 km²) to 6.1% (post-calibration at 3125 km²) for annual time periods. For monthly periods, the deviation of 34.9% (pre-calibration at 625 km²) was decreased to 15.4% (post-calibration at 3125 km²). Calibration can be improved further if more rain gauges are available. The GDA technique can be applied to calibrate TRMM rainfall data in regions with limited rain gauge data and can provide a sufficiently accurate estimate of the key hydrological process that can be used in water management applications.

Groundwater has a significant contribution to water storage and is considered to be one of the sources for agricultural irrigation; industrial; and domestic water use. The Gravity Recovery and Climate Experiment (GRACE) satellite provides a unique opportunity to evaluate terrestrial water storage (TWS) and groundwater storage (GWS) at a large spatial scale. However; the coarse resolution of GRACE limits its ability to investigate the water storage change at a small scale. It is; therefore; needed to improve the resolution of GRACE data at a spatial scale applicable for regional-level studies. In this study; a machine-learning-based downscaling random forest model (RFM) and artificial neural network (ANN) model were developed to downscale GRACE data (TWS and GWS) from 1° to a higher resolution (0.25°). The spatial maps of downscaled TWS and GWS were generated over the Indus basin irrigation system (IBIS). Variations in TWS of GRACE in combination with geospatial variables; including digital elevation model (DEM), slope; aspect; and hydrological variables; including soil moisture; evapotranspiration; rainfall; surface runoff; canopy water; and temperature; were used. The geospatial and hydrological variables could potentially contribute to; or correlate with; GRACE TWS. The RFM outperformed the ANN model and results show Pearson correlation coefficient (R) (0.97), root mean square error (RMSE) (11.83 mm), mean absolute error (MAE) (7.71 mm), and Nash–Sutcliffe efficiency (NSE) (0.94) while comparing with the training dataset from 2003 to 2016. These results indicate the suitability of RFM to downscale GRACE data at a regional scale. The downscaled GWS data were analyzed; and we observed that the region has lost GWS of about −9.54 ± 1.27 km³ at the rate of −0.68 ± 0.09 km³/year from 2003 to 2016. The validation results showed that R between downscaled GWS and observational wells GWS are 0.67 and 0.77 at seasonal and annual scales with a confidence level of 95%, respectively. It can; therefore; be concluded that the RFM has the potential to downscale GRACE data at a spatial scale suitable to predict GWS at regional scales.

The paper demonstrates the application of a new water accounting plus (WA+) framework to produce information on depletion of water resources, storage change, and land and water productivity in the Indus basin. It shows how satellite-derived estimates of land use, rainfall, evaporation (E), transpiration (T), interception (I) and biomass production can be used in addition to measured basin outflow, for water accounting with WA+. It is demonstrated how the accounting results can be interpreted to identify existing issues and examine solutions for the future. The results for one selected year (2007) showed that total annual water depletion in the basin (501 km³) plus outflows (21 km³) exceeded total precipitation (482 km³). The water storage systems that were effected are groundwater storage (30 km³), surface water storage (9 km³), and glaciers and snow storage (2 km³). Evapotranspiration of rainfall or "landscape ET" was 344 km³ (69 % of total depletion). "Incremental ET" due to utilized flow was 157 km³ (31% of total depletion). Agriculture depleted 297 km³, or 59% of the total depletion, of which 85% (254 km³) was through irrigated agriculture and the remaining 15% (44 km³) through rainfed systems. Due to excessive soil evaporation in agricultural areas, half of all water depletion in the basin was non-beneficial. Based on the results of this accounting exercise loss of storage, low beneficial depletion, and low land and water productivity were identified as the main water resources management issues. Future scenarios to address these issues were chosen and their impacts on the Indus Basin water accounts were tested using the new WA+ framework.