Estimating water balance components and their uncertainty bounds in highly groundwater-dependent and data-scarce area : An example for the Upper Citarum basin

Rusli, S.R.; Weerts, A.H.; Taufiq, A.; Bense, V.F.


Study region: Upper region of the Citarum basin in Indonesia. Study focus: Assessing water balance components in data-scarce regions using different approaches could result in different outcomes. In the upper reaches of the Citarum River in West Java, Indonesia, for example, many previous studies found the groundwater storage to be depleting, while GRACE identifies a contrasting trend of increasing groundwater storage change. Therefore, in this study, we aim to improve the accuracy of water balance components estimates in the Upper Citarum basin. Firstly, we estimate groundwater abstraction volumes based on population size and a review of literature. Estimates of the other water balance components, namely the rainfall, actual evaporation, discharge, and groundwater storage change are derived from various global datasets and available measurements. We also use a distributed hydrological model, wflow_sbm, to yield additional estimates of discharge and actual evaporation. We compare each basin water balance estimate and quantify the uncertainty of some components using the Extended Triple Collocation (ETC) method. New hydrological insights for the region: ETC application on four different rainfall estimates suggests a preference of using CHIRPS product in the study area as it delivers r2 of 0.56 and RMSE of 6.52 mm/day, compared to estimates from rainfall station (r2=0.39, RMSE = 8.57 mm/day), SACA&D (r2=0.29, RMSE = 10.46 mm/day), and TRMM (r2=0.56, RMSE = 8.61 mm/day). With CHIRPS rainfall forcing, wflow_sbm model estimates of average daily actual evaporation and discharge are obtained. The results for actual evaporation (2.67 mm/day) are plausible with a narrow difference of less than 0.50 mm/day among other estimates. The simulated discharge results in a daily average of 5.38 mm/day, estimated between observation data (3.65 mm/day) and GloFAS-ERA5 product (6.12 mm/day). Combining precipitation, actual evaporation, and discharge with a groundwater abstraction estimate of 0.57 mm/day, the wflow_sbm-based groundwater storage change is estimated at a daily storage depletion of 0.82 mm/day. Using GLEAM actual evaporation estimates (3.13 mm/day) and observed daily discharge, on the other hand, results in surplus water of 0.45 mm/day for groundwater storage change. These results demonstrate the high uncertainty in capturing subsurface hydrological processes, although the groundwater storage change estimates are found close to the estimates based upon GRACE gravimetric satellite data of 0.25 mm/day, with a variance of 1.57 mm/day. To aid in estimating current and future basin-scale groundwater level changes to support operational water management and policy in the Citarum basin, considering the massive groundwater abstraction, a focus on subsurface hydrological components quantification is of great importance for future research.