Quantifying total drainable water storage and its distribution using GRACE/GRACE-FO
Abstract
Total Drainable Water Storage (TDWS) is the portion of basin water storage that can be released through drainage and sustain river discharge. River discharge has intrinsic environmental value, as it provides the water needed to sustain river ecosystems, particularly during dry periods. A consistent global estimate of TDWS would therefore provide a new basis for comparing the capacity of river basins to sustain discharge and for identifying regions with comparatively limited drainable water-storage capacity. Despite its hydrological relevance, TDWS has not yet been estimated at the global scale from satellite gravimetry observations. As demonstrated in previous studies, TDWS can be inferred by linking temporal variations in basin water storage to river discharge through a runoff–storage relationship. GRACE and GRACE Follow-On (GRACE/GRACE-FO) provide a unique observational basis for this purpose by measuring Terrestrial Water Storage Anomalies (TWSA) across large river basins worldwide. However, TWSA includes storage components with different timing and routing characteristics relative to river discharge, which can introduce phase shifts and hysteresis into the storage–discharge relationship. We therefore derive Drainable Water Storage Anomalies (DWSA) by accounting separately for snow, lake, wetland, and reservoir storage. The resulting DWSA time series are combined with monthly river discharge observations from the Global Runoff Data Centre (GRDC) to characterize the runoff–storage relationship and estimate both TDWS and the characteristic drainage time, representing the timescale over which drainable storage is released through river discharge. Here, we provide the first global basin-scale quantification of TDWS from GRACE/GRACE-FO observations by applying this framework to more than 200 globally distributed river basins using three independent GRACE/GRACE-FO mascon solutions from CSR, JPL, and GSFC. Basin-scale TWSA and the corresponding DWSA are derived independently for each mascon product, enabling us to assess the consistency and robustness of TDWS and drainage-time estimates with respect to differences in gravity-field processing. We further examine the global spatial variability of both quantities and evaluate the suitability of the runoff–storage relationship for individual basins. This evaluation is important because uncertainties in river discharge, and weak or nonlinear storage–discharge behavior can affect the reliability of the inferred basin parameters. Together, TDWS and characteristic drainage time provide complementary information on how much water a basin can release through drainage and the timescale over which this storage is discharged. Their global estimation enables a consistent comparison of drainage behavior across river basins and helps identify differences in their capacity to sustain river flow during periods without precipitation.