Comparison between classic Delay-Doppler processing and a Fully-Focused SAR algorithm applied to a Sentinel-6 pass over the Itoiz reservoir. © isardSAT

Processing very high resolution satellite radar altimetry data using conventional algorithms takes several hours and even days, requiring multiple operations.

After developing a FF-SAR closed-form algorithm based on omega-K that substantially cuts down the execution time of Sentinel-6 radar altimetry data processing, we have designed an accelerated version of the the FFSAR back-projection algorithm that reduces the runtime of the classic algorithm.

Thanks to the accelerations of the algorithm and the implementation on a GPU-based architecture, this algorithm reduces up to 1,570 times the runtime of the classic algorithm running on a CPU.

​

Retrieving Earth’s geophysical parameters with Sentinel-6 data

The recent advances in signal processing algorithms have opened the door to new applications for altimetry data, such as the obtained by Sentinel-6 satellites.

Possibilities include iceberg detection, sea and land ice classification, ocean swell parameter estimation, the retrieval of key climate variables such as inland water level and extent, or coastal monitoring.

Another comparison between classic Delay-Doppler processing and a Fully-Focused SAR algorithm applied to a Sentinel-6 pass over the Greek island of Kato Koufonisi. © isardSAT.

​

Reference

S. Hernández-Burgos et al., “Accelerated Algorithm for Back-Projection Fully Focused SAR Altimetry,” in IEEE Transactions on Geoscience and Remote Sensing, vol. 63, pp. 1-16, 2025, Art no. 5209816, doi: 10.1109/TGRS.2025.3556544