Validation techniques for Radar altimeter measurements over ungauged inland waters in the Ethiopian highland and Rift valley
Dost, R.1; Maathuis, B.1; Ayana, E.2
1ITC; 2Bahirdar University
Altimeters are primarily designed to operate over uniform surfaces such as oceans and ice-sheets. Altimetry measurements over inland lakes can show deviations from in-situ gauge measurements. These deviations can be attributed to e.g. the geographically varying corrections applied to account for atmospheric effects on radar waves, and topographic variations along the shore line or within the lake (islands). To be able to use altimeter level data for lake management purposes, it is important that the data is validated. This paper will focus on this validation for inland lakes in the Ethiopian highland and Rift Valley, using Lake Tana as a test case. Since not all lakes in Ethiopia are gauged, an alternative RS supported validation technique for ungauged lakes was also tested.
First altimetry data obtained from ENVISAT and Topex/Poseidon (T/P) missions is compared with in-situ gauge data collected from 1992 - 2005. Then a bathymetric map is created of Lake Tana using a portable GPS linked sounder. By combining the bathymetry with SRTM data and optical RS a lake area/level relation could be established, that when compared with the in-situ gauge data could indicate suitability for radar altimetry validation.
The comparison between the in-situ level data and the Topex/Poseidon (T/P) and Envisat data shows a correlation of 0.97 and 0.76 respectively. The developed alternative remote sensing supported validation technique shows a correlation of 0.85.