CoAuthors

Christian Schwatke (DGFI-TUM, Germany); Denise Dettmering (DGFI-TUM, Germany)

The Sentinel-3A mission launched in 2016 continues the monitoring of inland water bodies. Sentinel-3A carries a Delay-Doppler/SAR altimeter that has a smaller along track footprint and improved range precision in comparison to classical LRM altimeters. In this work, we evaluate the performance of Sentinel-3A for observing heights of different inland water bodies, both rivers and lakes. Two different processing strategies for estimating water levels are tested: the DAHITI database processing and a classification approach with a k-means clustering algorithm based on features derived from the multi-look SAR waveform and the range integrated power (RIP) waveform. For the classification, the ESA provided L1B-S data set is used that includes the full stack matrix.

The derived water level time series are compared to each other and validated against in situ gauge data. In addition, the water level time series are compared to time series derived from Jason-3 observations. Jason-3 was also launched in 2016 but carries a conventional pulse limited altimeter. For the comparison, lakes observed by both missions are used. For rivers, locations have been selected with close distance (<20 km) between the virtual station of both missions and the in situ station.