Defining a retracking manifold within a radargram stack to improve satellite altimetric water level over coastal seas and inland water bodies

• Applicants: Professor Dr.-Ing. Nico Sneeuw; Dr.-Ing. Mohammad javad Tourian
• Subject Area: Geodesy, Photogrammetry, Remote Sensing, Geoinformatics, Cartography ; Hydrogeology, Hydrology, Limnology, Urban Water Management, Water Chemistry, Integrated Water Resources Management
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 496014904

Single-waveform retracking for satellite altimetry applications over coastal zones and inland waters has reached its limits, obtaining decimeter-level accuracy or worse. The existing retracking methods find a retracker offset in a waveform by analyzing the variation in backscattered power along the bin coordinate. This makes the retracking procedure strongly dependent on noise. Moreover, the success of such methods is only guaranteed for certain waveform types requiring cumbersome pre-processing steps including waveform classification. With the launch of the operational Sentinel-3 and 6 series of the European Copernicus programme, the algorithms to obtain highly precise water level estimates over inland waters and coastal seas need to become more robust, efficient and fit for automated use. Therefore, the main objective of this research project is to develop such a next-level retracking algorithm and, consequently, improve altimetric water level determination over inland waters and coastal regions. We aim to collect neighboring waveforms into a radargram and, then, to stack such radargrams over time. These so-called radargram stacks contain, unlike single waveforms, the full information on spatio-temporal variation of backscattered power over water surfaces. The radargram stack eases the recognition of patterns like retracking gate, off-nadir pattern (e.g. parabola), shoreline, etc. Instead of a retracking gate as a point in the 1D waveform, in a 3D radargram stack a surface referred to as retracking manifold is to be determined.In this research project, first the spatio-temporal evolution of satellite altimetry observations over water objects are analyzed and patterns in radargram stacks will be characterized. Then, in order to define the retracking manifold and its uncertainty, a Conditional Random Fields (CRF) will be developed. The CRF model benefits from conditional models in bin-space, bin-time and space-time slices. Then, a maximum a-posteriori solution will yield the retracking manifold. Then, the problem will be reformulated as an energy function minimization, for which the performance of different classes of optimization techniques is investigated. Finally, water level time series from retracked waveforms are validated against in situ data and will be compared with the performance of existing retracking algorithms. Within this study the altimetry data from three missions, Jason 2 &3 and Sentinel 3, are used. For the validation, several case studies with different climatological and hydrological conditions are selected.The idea of waveform retracking by analyzing its spatio-temporal behavior in a 3D data structure is formulated for the first time in this proposal and has not been addressed in the altimetry community before. It opens new pathways for achieving robust and more accurate water level estimates from operational missions and from future missions, e.g. SWOT, over inland waters and coastal seas.

• DFG Programme: Research Grants
• Co-Investigators: Dr.-Ing. Omid Elmi; Professor Dr. Uwe Sörgel