J6.3
Improving river discharge estimation by assimilating GRACE terrestrial water storage (TWS) retrievals into a distributed hydrological model: Water budget analysis in the Upper Zambezi River Basin (UZRB) and the Northern Kalahari Aquifer (NKA) in Southern Africa

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 4 February 2014: 9:00 AM
Room C209 (The Georgia World Congress Center )
Jing Tao, Duke University, Durham, NC; and A. P. Barros

Hydrological modeling is generally limited by the accuracy of forcing datasets, and the availability of ancillary data to support appropriate physical representation of hydrological processes. In the Upper Zambezi River Basin (UZRB) in Southern Africa, this is an especially critical challenge as there are virtually no ground-based hydrologic or hydrometeorological measurements to support modeling activities much less predictive studies. The terrestrial water storage (TWS) observations by NASA's Gravity Recovery and Climate Experiment (GRACE) satellite provide a unique opportunity to evaluate and constrain basin-scale hydrologic models at large spatial scales, which is the case of the UZRB. In particular, through Data Assimilation techniques to optimally merge satellite observations and hydrological models, there is an opportunity to investigate hydrological processes and begin to develop a quantitative understanding of the regional water cycle in what is arguably the most critical river basin in Southern Africa. In this study, we present a Hydrologic Data Assimilation System (HDAS) which relies on a three-dimensional coupled surface-groundwater hydrology model (3D-LSHM) integrated with an EnKS-based (Ensemble Kalman Smoother) data-assimilation system, to characterize the seasonal (wet/dry season) and inter-annual variability of the water budget of the UZRB and the underlying Northern Kalahari Aquifer (NKA) at high temporal-spatial resolution for the ten-year period 2003-2012 of GRACE TWS L3 RL05 data. The EnKS, which is at least as good as EnKF (Ensemble Kalman Filter) since it takes the EnKF estimate as the first guess, is adopted mainly due to deal with temporal scaling issues that result from the relatively coarse monthly scale of GRACE data. The control variables of the assimilation scheme include soil moisture of each layer in the vadose zone, and the depth of water table. The absolute TWS is constructed by adding the mean of ten-year open-loop simulation to the GRACE TWS anomalies data similar to Zaitchik et al.(2008) though a fully distributed hydrology model is used here. Transboundary groundwater fluxes are neglected. Preliminary results demonstrate that the river discharge estimation is improved by assimilating the GRACE-derived TWS into the 3D-LSHM. The modeling system subsequently can be used in prognostic mode to assess the impact of LULC and climate change scenarios on groundwater resources vulnerability in the Upper Zambezi.

Zaitchik, B.F., Rodell, M., Reichle, R.H., 2008. Assimilation of GRACE terrestrial water storage data into a Land Surface Model: Results for the Mississippi River basin. J. Hydrometeorol., 9(3): 535-548.