Assimilating GRACE terrestrial water storage data into a conceptual hydrology model for the River Rhine
Terrestrial water storage refers to all water stored on the land surface including surface water, soil moisture, groundwater, vegetation water content, snow, ice and permafrost. Terrestrial water storage is a key component of the terrestrial and global hydrological cycles, and plays a major role in the Earth’s climate.
The Gravity Recovery and Climate Experiment (GRACE) twin satellite mission provided the first space-based dataset of TWS variations. In this study, we want to determine the value of assimilating GRACE observations into a well-calibrated conceptual hydrology model of the Rhine river basin. We are using two data assimilation methods (the ensemble Kalman filter (EnKF) and smoother (EnKS)) to assimilate the GRACE TWS variation data into the HBV-96 rainfall run-off model.
GRACE Twin Satellites (Image from CSR )
There are many GRACE products, developed at different research centers, each using their own data processing and filtering methods to retrieve TWS from the GRACE signal. One of the first things we need to do is compare the different GRACE products to see how similar they are in the Rhine river basin. Data assimilation usually yields an estimate somewhere between the modeled estimate and the observation, so any differences between the products will influence our results. We use several different GRACE products including the TU Delft DMT-1, the CSR-RL04 from the University of Texas at Austin, and a TU Delft product derived using radial basis functions.
Our results are validated against streamflow at three locations. If the estimated streamflow improves after assimilation is closer to the observed streamflow, then assimilation led to improved model performance. This is an ongoing project in which we are continually improving the data assimilation strategy.
If you would like to do an MSc. thesis on this topic, please contact Susan Steele-Dunne