The aim of this study is the improvement of the TanDEM-X elevation model for future floodwater modeling by implementing surveyed road dams and the use of filter algorithms. Modern satellite systems like TanDEM-X deliver high-resolution images with a high vertical and horizontal accuracy. Nevertheless, regarding special usage they sometimes reach their limits in documenting important features that are smaller than the grid size. Especially in the context of 2D-hydrodynamic flood modelling, the features that influence the runoff processes, e.g. road dams and culverts, have to be included for precise calculations. To fulfil the objective, the main road dams were surveyed, especially those that are blocking the flood water flowing from south Angola to the Etosha Pan in northern Namibia. First, a Leica GS 16 Sensor was installed on the roof of a car recording position data in real time while driving on the road dams in the Cuvelai Basin. In total, 532 km of road dams have been investigated during 4 days while driving at a top speed of 80 km/h. Due to the long driving distances, the daily regular adjustment of the base station would have been necessary but logistically not possible. Moreover, the lack of reference stations made a RTK and Network-RTK solution likewise impossible. For that reasons, the Leica SmartLink function was used. This method is not dependent on classic reference stations next to the GNSS sensor but instead works with geostationary satellites sending correction data in real time. The surveyed road dam elevation data have a vertical accuracy of 4.3 cm up to 10 cm. These precise measurements contribute to rectifying the TanDEM-X elevation data and thus improve the surface runoff network for the future floodwater model and should enhance the floodwater prediction for the Cuvelai Basin.
