Abstract's details
Calibration of Sentinel-3 altimeter data using a corner reflector
CoAuthors
Event: 2023 Ocean Surface Topography Science Team Meeting
Session: Regional and Global CAL/VAL for Assembling a Climate Data Record
Presentation type: Type Poster
Contribution: PDF file
Abstract:
Corner reflectors (CR) are commonly used to calibrate and characterize the quality of SAR images. With the Fully Focused SAR (FFSAR) method, the along-track resolution is now sufficient for CR to be detected by altimeters, opening new possibility for altimetry calibration with CR.
CR are passive devices (by opposition to the transponder active devices historically used for altimetry calibration). In comparison to the transponders, this offers some advantages: the development is relatively inexpensive, it has a very broadband frequency response, maintenance is reduced, the deformation of the across-track impulse response, a known issue with the transponders, is not present on CR, there is no need in regular calibration as it is the case for transponders’ internal path delay… Also, because the signal is only backscattered and not amplified, CR theoretically allow for the calibration and the monitoring of the absolute sigma0 if the noise signal ratio of the site is sufficient.
For all these reasons, CNES has developed a CR in 2019. The 2 m squared side CR was deployed on 8th October 2020 in Toulouse under a Sentinel-3 track. The place has been chosen for its particular topography, and its sufficiently low clutter level. A GPS positioning campaign has been performed to locate precisely the CR. The on-board OLTC has been modified to update the tracking window to target the CR theoretical range distance.
In this study we present the analysis of Sentinel-3 CR acquisitions. Range bias, datation bias, and sigma0 have been estimated from FFSAR processing (2D PTR method). Range bias has been corrected from geophysical corrections extracted from the operational products, and from the distance between the GNSS acquisition and the CR phase center position. Sigma0 has been corrected from gain and attenuation correction taken from the level-2 operational products; the antenna pattern has also been compensated. Impulse response has also been used to monitor range and azimuth resolution. For sake of validation, the results have been compared with CDN1 transponder device at the same cycles and the same method has been used on both CR and transponder. Analysis show very good consistency between CR and transponder results in terms of bias and stability for range and datation bias. The sigma0 stability is also discussed.
Detections of the CR with other sensors working in different bands are also shown: detection in Sentinel-1 C-band SAR images during the campaign under the Sentinel-3 track, and, more recently, detection under the SWOT Ka-band swath.
CR are passive devices (by opposition to the transponder active devices historically used for altimetry calibration). In comparison to the transponders, this offers some advantages: the development is relatively inexpensive, it has a very broadband frequency response, maintenance is reduced, the deformation of the across-track impulse response, a known issue with the transponders, is not present on CR, there is no need in regular calibration as it is the case for transponders’ internal path delay… Also, because the signal is only backscattered and not amplified, CR theoretically allow for the calibration and the monitoring of the absolute sigma0 if the noise signal ratio of the site is sufficient.
For all these reasons, CNES has developed a CR in 2019. The 2 m squared side CR was deployed on 8th October 2020 in Toulouse under a Sentinel-3 track. The place has been chosen for its particular topography, and its sufficiently low clutter level. A GPS positioning campaign has been performed to locate precisely the CR. The on-board OLTC has been modified to update the tracking window to target the CR theoretical range distance.
In this study we present the analysis of Sentinel-3 CR acquisitions. Range bias, datation bias, and sigma0 have been estimated from FFSAR processing (2D PTR method). Range bias has been corrected from geophysical corrections extracted from the operational products, and from the distance between the GNSS acquisition and the CR phase center position. Sigma0 has been corrected from gain and attenuation correction taken from the level-2 operational products; the antenna pattern has also been compensated. Impulse response has also been used to monitor range and azimuth resolution. For sake of validation, the results have been compared with CDN1 transponder device at the same cycles and the same method has been used on both CR and transponder. Analysis show very good consistency between CR and transponder results in terms of bias and stability for range and datation bias. The sigma0 stability is also discussed.
Detections of the CR with other sensors working in different bands are also shown: detection in Sentinel-1 C-band SAR images during the campaign under the Sentinel-3 track, and, more recently, detection under the SWOT Ka-band swath.