|Developing an Ensemble Kalman Filter calibration and data assimilation (EnC/DA) approach for integrating geodetic and remote sensing data into a global hydrological model|
In project P1, we will integrate WaterGAP Global Hydrological Model (WGHM, see Project 2) simulations of terrestrial water fluxes and storages with geodetic and remote sensing data sets through data assimilation (DA). While several techniques exist for data assimilation, here we will utilize the Ensemble Kalman Filter/Smoother approach (EnKF/KS). One of the reasons for this is that it is relatively easy to estimate model parameter in the EnKF/KS, something which is termed model calibration in hydrology.
Figure 1: Combining GRACE data and the WGHM
(generated by Kerstin Schulze, University of Bonn)
From a hydrological point of view, DA will provide a more realistic view on water storages than model simulations allow, and it helps to identify shortcomings in model structure and climate forcing data. From a geodetic point of view, DA enables one to disaggregate GRACE and other data sets like e.g. satellite-derived water levels vertically, horizontally and temporally. For example, Fig. 2 demonstrates the downscaling of GRACE through DA. In addition, DA nudges the model closer to the observations.
Figure 2: Assimilation of GRACE-derived TWSA into the WGHM for the Murray-Darling-Basin in Australia
(generated by Helena Gerdener, University of Bonn, and Maike Schumacher, University of Hohenheim)
On this website, we provide some basic information about developing a DA framework for the Global CDA research unit and the different data sets that will be assimilated into the WGHM model.