Time and Frequency Domain Numerical SAR Retracking for Sentinel-6: First results prior to the S6PP implementation

Salvatore Dinardo (CLS, France)


Thomas Moreau (CLS, FRANCE); Claire Maraldi (CNES, FRANCE); Francois Boy (CNES, FRANCE); nicolas picot (CNES, FRANCE)

Event: 2020 Ocean Surface Topography Science Team Meeting (virtual)

Session: Instrument Processing: Measurement and Retracking

Presentation type: Type Forum

In the frame of the exploitation of the Sentinel-6 altimetry mission, CNES has contracted CLS for the development of the Sentinel-6 Processing Prototype (S6PP). S6PP is a multi-chain (LRM, LR-RMC, UF-SAR, FF-SAR) processor in which the novel algorithms developed in the CNES/CLS R&D activities are implemented and validated in support to the different thematic applications (in particular inland water and ocean) and in view of promoting their possible implementation in operational ground segment.
The present work covers in particular the unfocused SAR (UF-SAR) chain of the S6PP detailing all the novelties which are planned to be implemented in this branch at L1B and L2 and which represents a processing complementary to what is going to be carried out in the operational PDAP context.
S6PP will have a strong heritage from CLS SMAP processor (Rieu at al., 2020).
In particular, as far as concerns the UF-SAR chain, at L1b it is proposed the implementation of a standard Delay-Doppler processing complemented by the Range Walk Correction application (Scagliola et al. 2019) and by an exact beam-steering and a beam formation by DFT (Discrete Fourier transform) or CZT (Chirp Zeta transform) for sake of CPU time saving.
At L2, in order to be less prone to errors in range coming from the range PTR shape distortions along the mission, a real-PTR based Numerical SAR Retracking is proposed in frequency-domain and in the time-domain, both having as minimization problem solver the Levenberg–Marquardt algorithm.
The frequency-domain approach will be based on the computation of a waveform model in the frequency domain as proposed by Buchhaupt et al. (2017) whereas the time-domain will be based on the resolution of the SAR waveform equation in the time domain (S. Dinardo et al. in prep.).
The first approach will be the one used for processing SAR data routinely over ocean, also providing less operational constraint, whereas the second one will be just used to tackle dedicated analysis on specific and detailed points as for instance impact of the real antenna pattern on the data accuracy/precision.
Both approaches will modelize natively the azimuth ambiguities in the Delay-Doppler Map model and hence will not require any ambiguity masking.
First results outlining the impact of the range walk for Sentinel-6, the impact of the Beam sub-sampling (S. Figerou et al. at OSTST-2020) and the good consistency between the results from SAR numerical retracking (time domain) versus the S6 GPP data results will be shown during the presentation.
We acknowledge ESA and EUMETSAT for the provision of S6 test datasets necessary to carry out this work and CNES for the expertise and funding support.


Buchhaupt C., Fenoglio-Marc L., Dinardo S., Scharroo R., Becker M (2017). A fast convolution based waveform model for conventional and unfocused SAR altimetry, Advanced Space Research Special Issue CryoSat-2, https://doi.org/10.1016/j.asr.2017.11.039

Rieu P., Moreau T., Cadier E., Raynal M., Clerc S., Donlon C., Borde F., Boy F., Maraldi C. (2020): Exploiting the Sentinel-3 tandem phase dataset and azimuth oversampling to better characterize the sensitivity of SAR altimeter sea surface height to long ocean waves, Advances Space Research, https://doi.org/10.1016/j.asr.2020.09.037

Scagliola, M., Recchia, L., Maestri, L., Giudici, D., Evaluating the impact of range walk compensation in delay/Doppler processing over open ocean, Advances in Space Research (2019), doi: https://doi.org/10.1016/j.asr.2019.11.032
Salvatore Dinardo