Renou et al. (2025) Innovative validation of Sentinel-3 SAR altimetry measurements over rivers for the evaluation of hydrology-dedicated retrackers

Identification

• Journal: Advances in Space Research
• Year: 2025
• Date: 2025-11-11
• Authors: Julien Renou, Marie Chapellier, Jérémie Aublanc, Nicolas Taburet, Marco Restano, Filomena Catapano, Alessandro Di Bella, Pierre Féménias
• DOI: 10.1016/j.asr.2025.11.005

Research Groups

• Collecte Localisation Satellites (CLS), Toulouse, France
• Starion Group c/o ESA-ESRIN, Frascati, Italy
• ESA-ESRIN, Frascati, Italy

Short Summary

This study introduces an innovative "off-nadir" validation method for Sentinel-3 SAR altimetry over rivers, which significantly reduces the impact of river slope on Water Surface Height (WSH) estimates. Applying this method to French rivers, the research statistically evaluates hydrology-dedicated retrackers, finding a WSH bias of approximately 0.04 m for the sinc2 retracker (reducible to near zero in favorable conditions) and 0.22 m for the OCOG retracker.

Objective

• To develop and apply an innovative "off-nadir" validation method for Sentinel-3 SAR altimetry measurements over rivers to reduce the impact of river slope.
• To statistically evaluate the bias of hydrology-dedicated retracking algorithms (OCOG and sinc2) using real altimetric measurements over isolated narrow rivers.
• To investigate the impact of river slope for both nadir and off-nadir validation techniques and identify favorable scenarios for the off-nadir approach.

Study Configuration

• Spatial Scale: French continental territory, focusing on isolated narrow rivers (most under 100 m wide). In situ stations located within 2000 m of Sentinel-3 theoretical ground tracks. Sentinel-3 SAR band along-track footprint is approximately 400 m, and across-track footprint is 15000 m. River sections sampled within the SAR band ranged from 800 m to 2000 m.
• Temporal Scale: Data from Sentinel-3A (launched February 2016) and Sentinel-3B (launched April 2018) missions. In situ records predate Sentinel-3A. SWOT mission data from March to November 2024 was used for independent river slope validation. A maximum time span of 4 hours was allowed between Sentinel-3 and in situ measurements.

Methodology and Data

• Models used:
  • Retracking algorithms: Offset Centre Of Gravity (OCOG/ICE-1) and sinc2 retracker.
  • Waveform classification algorithm (neural network-based) from Poisson et al. (2018) adapted for SAR altimetry (Longepe et al., 2019).
  • Levenberg–Marquardt optimization algorithm for sinc2 function fitting.
• Data sources:
  • Satellite altimetry: ESA Sentinel-3 Level-2 hydrology thematic products (20 Hz UnFocused-Synthetic Aperture Radar (UF-SAR) measurements from Sentinel-3A and Sentinel-3B SRAL instrument).
  • In situ measurements: French SCHAPI network (hundreds of stations providing WSH timeseries with centimeter-level precision and 15-minute sampling).
  • Geoid reference: IGN1969 geoid (converted to WGS-84 ellipsoid using the French National Institute of Geographic and Forest Information (IGN) grid RAF20.tac).
  • Water mask: Topage database (high-resolution, used for river centerline definition and contaminant detection).
  • SWOT mission: PIXel Cloud (PIXC) product (used for independent river slope estimation).

Main Results

• The "off-nadir" validation method significantly reduced the impact of river slope, improving precision. For a single station, the Median Absolute Deviation (MAD) decreased from 0.469 m to 0.057 m, and Root Mean Square Error (RMSE) from 1.205 m to 0.482 m (using OCOG).
• Statistically, the off-nadir validation increased the ratio of Sentinel-3 WSH values with a small deviation from in situ WSH values by a factor of 2 compared to nadir validation. Over 90% of off-nadir measurements had an absolute WSH difference less than 1.2 m, compared to less than 55% for nadir measurements.
• For the OCOG retracker, a median WSH bias of approximately 0.22 m was found.
• For the sinc2 retracker, a median WSH bias of approximately 0.04 m was found.
• When selecting favorable scenarios (river sections sampled within the SAR band not exceeding 800 m), the residual bias of the sinc2 retracker reached a value close to zero, while the OCOG retracker's bias remained around 0.18 m.
• The study demonstrated that off-nadir validation applied to UF-SAR Sentinel-3 measurements can retrieve time-averaged river slopes, consistent with independent SWOT mission estimates (e.g., -0.00182 m/m from Sentinel-3 vs. -0.00179 m/m from SWOT for the Dordogne river).

Contributions

• Introduction of an innovative "off-nadir" validation method for SAR altimetry over rivers, which effectively reduces the impact of river slope on WSH estimates, addressing a major limitation of standard nadir validation.
• First statistical characterization of the OCOG retracker bias (approximately 0.22 m) over rivers using real Sentinel-3 altimetric measurements.
• Demonstration of the superior performance of the sinc2 retracker over OCOG for isolated narrow rivers, showing a significantly lower bias (approximately 0.04 m, reducible to near zero in favorable conditions).
• Identification of "favorable scenarios" (rivers flowing predominantly along-track, short sampled river sections) where the off-nadir validation yields near-zero bias for sinc2, enabling more rigorous retracker evaluation.
• Demonstration of the capability of UF-SAR Sentinel-3 off-nadir measurements to retrieve time-averaged river slopes, validated against SWOT data.
• Highlights the potential for off-nadir estimates to support hydrological applications (calibrating models, evaluating discharge, rating curves) and investigate ungauged basins, especially in areas with high-resolution water masks.

Funding

• ESA/ESRIN, through the "Sentinel-3 Mission Performance Cluster" project (ESA contract number 4000136824/21/I-BG).

Citation

@article{Renou2025Innovative,
  author = {Renou, Julien and Chapellier, Marie and Aublanc, Jérémie and Taburet, Nicolas and Restano, Marco and Catapano, Filomena and Bella, Alessandro Di and Féménias, Pierre},
  title = {Innovative validation of Sentinel-3 SAR altimetry measurements over rivers for the evaluation of hydrology-dedicated retrackers},
  journal = {Advances in Space Research},
  year = {2025},
  doi = {10.1016/j.asr.2025.11.005},
  url = {https://doi.org/10.1016/j.asr.2025.11.005}
}