Data Assimilation for unsaturated flow in heterogenous soils
M.Sc. Natascha Brandhorst, Prof. Dr. Insa Neuweiler
Funding: DFG research unit FOR2131
Flow in the unsaturated zone is crucial for the hydrological cycle as it influences some of its main components, i.e. groundwater recharge or root water uptake. Its governing equations are highly non-linear and depend on various parameters that are difficult to estimate. In small scale models, calibration can help determine these parameters that enable the model to make reasonable predictions of water content, groundwater recharge or similar. In large scale models, calibration becomes computationally very demanding. In contrast to calibration methods, Data assimilation can, to a certain degree, offer good predictions and parameter estimates at much smaller computing time making it attractive for large models.
On large scale, soils are characterized by heterogeneities, though. Resolving these heterogeneous structures would counteract the speed-up gained by the efficient data assimilation method. Furthermore, the heterogeneities can hardly be captured by sparse local measurements. Therefore, the aim is to find a model of lower complexity that is able to represent the mean behavior of the large scale heterogeneous reality and predict the fluxes entering or leaving this compartment. To that purpose the impact on the quality of water content prediction of different heterogeneous structures, roots penetrating the vadose zone, location of measurement devices, as well as different data assimilation methods will be investigated.
Letzte Änderung: 03.11.23
