Abstract's details
Analysis of particular aspects of the Sentinel-6 altimeter processing using Cryosat-2 data
CoAuthors
Event: 2016 SAR Altimetry Workshop
Session: Future missions, recommendations and round table
Presentation type: Type Poster
Contribution: not provided
Abstract:
The Jason-CS/Sentinel-6 mission is an operational oceanography program of two satellites, flying in sequence, which will continue the successful Jason series of ocean-monitoring satellites over at least a decade from 2020 to beyond 2030. The Jason-CS satellites will carry a radar altimeter, the Poseidon-4 instrument, evolving significantly from its predecessors, the Poseidon-3A and -3B instruments of Jason-2 and -3 missions, and the Sentinel-3 SRAL.
Poseidon-4 is a dual frequency (Ku-band and C-band) radar altimeter with an innovative interleaved chronogram at 9-kHz Pulse Rate Frequency (PRF). This low PRF allows pulses reception between two consecutive emissions without any breaks or long waiting period, making it possible to continuously emit signals and in turn achieve a better exploitation of the time of measurements compared to the closed-burst operational mode of Cryosat-2 and Sentinel-3 (where 70% of time is not used). That way, the high-rate interleaved pulsing will theoretically further enhance noise reduction (by a factor of 1.7 compared to Sentinel-3 SAR mode).
However, this PRF is not high enough to adequately sample the total Doppler frequency band induced by the motion of the satellite (~14 kHz). The choice of a PRF that does not satisfy the Nyquist sampling criterion will lead to the appearance of Doppler ambiguities which will modify the shape of the SAR echoes and may alter the estimates obtained if the usual ocean SAR retrackers implemented in CryoSat-2 and Sentinel-3 were used.
At this time, very few studies have actually examined the impact of the Doppler ambiguities. And so far, these analyses have only been performed on simulated data, generated either by a numerical model of the Sentinel-6 radar altimeter echoes or by means of the mission performance simulator. The results of these studies concluded that the aliasing has either to be accounted in the SAR/Doppler model or to be cut out to ensure good consistency of the estimates. Such computer simulations allowed to reveal a theoretical understanding of the innovative SAR interleaved method but this analysis has not yet been demonstrated with real data, over different sea state conditions. It is actually essential to undertake investigations based on the use of real data to better characterize and understand the effects of these ambiguities, and find any potential anomalies in SAR altimetry data or in the sea level content that would be caused by these ambiguities.
This paper presents the main results of a study that aimed at thoroughly analyzing the effects of lowering the PRF, on SAR-mode estimates, using Cryosat-2 data processed at the PRF of the Sentinel-6 altimeter over ocean (1 pulse out of 2 is kept in the 18 kHz Cryosat-2/SIRAL bursts of 64 pulses to produce SAR altimeter data with Doppler aliasing artifacts). Different models have been investigated (accounting or not for the aliasing, filtering out the ambiguities) and the delay-Doppler processor adapted accordingly. A full assessment has been performed for each data set in global and over large time period (at least 1 month of data) and compared to the Cryosat-2 SIRAL reference.
Second, this paper deals with the analysis of an experimental LR mode that will be exploited by the Sentinel-6 mission. The so-called LR-RMC mode designed by TAS [Phalippou and Demeestere, OSTST 2011] differs from the classical SAR-mode processing in a number of ways. In the same manner as the SAR-mode processing, this mode allows processing data coherently dividing the radar footprint into contiguous strips of 300 m width. However, contrary to the SAR-mode processing, the LR-RMC mode does not stack and average incoherently looks at a same surface sample. After azimuthally processing, the formed Doppler beams from the same burst are range corrected and then directly summed with each other. The resulting LR-RMC echoes have a similar peaky shape, but a degraded spatial resolution (comparable to the LRM one).
This new processing approach has been recently studied with Cryosat-2 data showing potential improvement of the SAR altimetric measurements’ precision over ocean, at the expense of a lower spatial resolution. A data set of at least one-month duration has been produced over open ocean and their performance assessed through the use of common validation tools and protocols via comparison with operational-like data products (generated with CNES processor CPP). The aim is to determine whether this new processing scheme has a potential interest in operational use or not.
Poseidon-4 is a dual frequency (Ku-band and C-band) radar altimeter with an innovative interleaved chronogram at 9-kHz Pulse Rate Frequency (PRF). This low PRF allows pulses reception between two consecutive emissions without any breaks or long waiting period, making it possible to continuously emit signals and in turn achieve a better exploitation of the time of measurements compared to the closed-burst operational mode of Cryosat-2 and Sentinel-3 (where 70% of time is not used). That way, the high-rate interleaved pulsing will theoretically further enhance noise reduction (by a factor of 1.7 compared to Sentinel-3 SAR mode).
However, this PRF is not high enough to adequately sample the total Doppler frequency band induced by the motion of the satellite (~14 kHz). The choice of a PRF that does not satisfy the Nyquist sampling criterion will lead to the appearance of Doppler ambiguities which will modify the shape of the SAR echoes and may alter the estimates obtained if the usual ocean SAR retrackers implemented in CryoSat-2 and Sentinel-3 were used.
At this time, very few studies have actually examined the impact of the Doppler ambiguities. And so far, these analyses have only been performed on simulated data, generated either by a numerical model of the Sentinel-6 radar altimeter echoes or by means of the mission performance simulator. The results of these studies concluded that the aliasing has either to be accounted in the SAR/Doppler model or to be cut out to ensure good consistency of the estimates. Such computer simulations allowed to reveal a theoretical understanding of the innovative SAR interleaved method but this analysis has not yet been demonstrated with real data, over different sea state conditions. It is actually essential to undertake investigations based on the use of real data to better characterize and understand the effects of these ambiguities, and find any potential anomalies in SAR altimetry data or in the sea level content that would be caused by these ambiguities.
This paper presents the main results of a study that aimed at thoroughly analyzing the effects of lowering the PRF, on SAR-mode estimates, using Cryosat-2 data processed at the PRF of the Sentinel-6 altimeter over ocean (1 pulse out of 2 is kept in the 18 kHz Cryosat-2/SIRAL bursts of 64 pulses to produce SAR altimeter data with Doppler aliasing artifacts). Different models have been investigated (accounting or not for the aliasing, filtering out the ambiguities) and the delay-Doppler processor adapted accordingly. A full assessment has been performed for each data set in global and over large time period (at least 1 month of data) and compared to the Cryosat-2 SIRAL reference.
Second, this paper deals with the analysis of an experimental LR mode that will be exploited by the Sentinel-6 mission. The so-called LR-RMC mode designed by TAS [Phalippou and Demeestere, OSTST 2011] differs from the classical SAR-mode processing in a number of ways. In the same manner as the SAR-mode processing, this mode allows processing data coherently dividing the radar footprint into contiguous strips of 300 m width. However, contrary to the SAR-mode processing, the LR-RMC mode does not stack and average incoherently looks at a same surface sample. After azimuthally processing, the formed Doppler beams from the same burst are range corrected and then directly summed with each other. The resulting LR-RMC echoes have a similar peaky shape, but a degraded spatial resolution (comparable to the LRM one).
This new processing approach has been recently studied with Cryosat-2 data showing potential improvement of the SAR altimetric measurements’ precision over ocean, at the expense of a lower spatial resolution. A data set of at least one-month duration has been produced over open ocean and their performance assessed through the use of common validation tools and protocols via comparison with operational-like data products (generated with CNES processor CPP). The aim is to determine whether this new processing scheme has a potential interest in operational use or not.