Multi-peak Retracking of CryoSat-2 SARIn Waveforms Over Arctic Sea Ice

• Published in: IEEE Transactions on Geoscience and Remote Sensing
• Main Authors: Di Bella, Alessandro, Kwok, Ronald, Armitage, Thomas W. K., Skourup, Henriette, Forsberg, René
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2021
• Subjects: Sea ice; Radar altimetry; Retracking; CryoSat-2; Interferometry; Swath processing; Arctic
• Online Access: https://orbit.dtu.dk/en/publications/ad08e68c-14d1-44df-95bf-ac58df197faf; https://doi.org/10.1109/TGRS.2020.3022522; https://backend.orbit.dtu.dk/ws/files/221196775/09205871.pdf; https://backend.orbit.dtu.dk/ws/files/245797984/09205871.pdf

CryoSat-2 (CS2) is the first mission equipped with a pulse-limited radar altimeter capable of operating in Synthetic Aperture Radar (SAR) Interferometric (SARIn) mode. Over ice sheets and ice caps, CS2 SARIn data have been used to retrieve surface elevations over an across-track ground ``swath.'' This work demonstrates that retracking multiple coherent peaks of CS2 SARIn waveforms, in combination with the interferometric phase, enables to obtain more than one valid height estimate from single SARIn waveforms over Arctic sea ice. For some SARIn waveforms, the scattering from sea ice at the satellite nadir is successfully separated from returns originating from off-nadir leads. An average bias of -1.8 cm is found for absolute sea ice elevations when using a 50% threshold retracker. It is shown that including multiple SARIn peaks and the associated phase difference in the processing does not introduce any bias on the average sea ice freeboard heights compared with the estimates from regular SAR processing schemes, while significantly increasing the number of valid sea surface height retrievals (+55%) and the number of freeboard estimates in the coastal domain and in multi-year ice regions (~3 times). This results in an average ~34% reduction of the gridded random freeboard uncertainty, corresponding to a ~20% reduction of the gridded total sea ice thickness uncertainty. The results of this work show that SARIn acquisitions over Arctic sea ice provide improved spatial coverage and denser sampling of sea level and sea ice freeboard compared with the SAR mode, with accuracy being largely driven by the retracking algorithm.