About

Log in?

DTU users get better search results including licensed content and discounts on order fees.

Anyone can log in and get personalized features such as favorites, tags and feeds.

Log in as DTU user Log in as non-DTU user No thanks

DTU Findit

Journal article

Evaluation of Sentinel-3 SRAL SAR altimetry over Chinese rivers

From

Department of Environmental Engineering, Technical University of Denmark1

Air, Land & Water Resources, Department of Environmental Engineering, Technical University of Denmark2

National Space Institute, Technical University of Denmark3

Geodesy and Earth Observation, National Space Institute, Technical University of Denmark4

HE Space Operations GmbH5

Satellite radar altimetry observations of water surface elevation (WSE) have become an important data source to supplement river gauge records. Sentinel-3 is the first radar altimetry mission operating with a synthetic aperture radar (SAR) altimeter at global scale and with a new on-board tracking system (i.e. open-loop), which has great potential in terms of delivering reliable observations of inland water bodies for the next two decades (several future missions include an open-loop tracking mode).

In this context, it is very important to investigate the data quality at an early stage. In this study, a comprehensive evaluation of Sentinel-3A (S3A) is conducted at 50 virtual stations (VS) located on a wide range of rivers in China. The evaluation of Level 1 data shows that, over mountain rivers, a good prior surface elevation estimate on-board is vital to deliver useful datasets using the S3A open-loop tracking system.

The Open-Loop Tracking Command version 5 (OLTC V5) has significantly improved the placement of the range window, which was misplaced and resulted in lack of data over many mountain rivers prior to OLTC V5 (March 2019). However, application of S3A over mountain rivers still requires careful evaluation, especially before March 2019.

Four retrackers are evaluated including a physical SAR Altimetry Mode Studies and Applications retracker (SAMOSA+), a traditional Offset Center Of Gravity (OCOG), a Primary Peak Center Of Gravity (PPCOG), and a modified Multiple Waveform Persistent Peak (MWaPP+) retracker. For 26 VSs in plain areas, retracked WSE data achieved a root mean square error (RMSE) ranging from 0.12 m to 0.9 m.

The comparison of retracking methods reveals that SAMOSA+, OCOG, and PPCOG are unable to handle multi-peak waveforms. But the MWaPP+ can significantly improve the accuracy of the estimated WSE over large rivers, especially when the waveforms are contaminated. Moreover, our result shows no considerable difference between medium (ca. 300 m wide) and large (wider than 500 m) rivers.

Instead, surrounding topography and homogeneity of surroundings are very important factors influencing the shape of a waveform. For rivers surrounded by lakes, man-made channels etc., special care must be taken when processing altimetry data. Dedicated retracking methods, such as MWaPP+, and sophisticated methods for outlier detection are needed to improve precision over such rivers, as demonstrated here for the Yangtze River.

Language: English
Year: 2020
Pages: 111546
ISSN: 18790704 and 00344257
Types: Journal article
DOI: 10.1016/j.rse.2019.111546
ORCIDs: Jiang, Liguang , Nielsen, Karina , Andersen, Ole Baltazar and Bauer-Gottwein, Peter

DTU users get better search results including licensed content and discounts on order fees.

Log in as DTU user

Access

Analysis