Monitoring inland surface water level from Sentinel-3 data

Abstract

Inland surface water bodies (e.g. lakes and rivers) are very important to the nature and human society. To monitor the water level of inland water bodies, gauge stations were built since 19th century, but the amount of the stations is declining since the 1970s because of lack of maintenance. An accurate and continuous monitoring of lakes and rivers is available because of the satellite altimetry missions launched, e.g. Jason-2 and ENVISAT. These satellites can provide water level with proper spatial and temporal resolution.

In the recent past, researchers have used different satellite mission observations to generate time series of inland water level in order for monitoring the water bodies. In this thesis, we use the new designed satellite mission Sentinel-3, which carries different sensors, to generate the water level time series of Dongting Lake and Poyang Lake in China. Initially, we combine the altimetry measurements with satellite images to determine virtual station. We choose Sentinel-3 Ku band data and on-board Ocean tracker to generate the water level time series. Afterwards, we apply different waveform retracking algorithms (5β-parameter and OCOG) to compare the results with on-board tracker. We also validate the results with the other database, then investigate the waveforms of each sampling date.

The comparisons show the three tracking methods we used are capable to Quasi-Specular waveforms, and OCOG shows the best result to flat patch waveforms. Furthermore, some suggestions for improvements are also discussed in the last chapter.