MulPOS-C: A Spatiotemporal Continuity-Constrained Multi-Parameter Sub-Waveform Retracker for Coastal Altimetry

Coastal zones are closely associated with human activities and socio-economic development. Satellite altimetry is a key technology for coastal sea-level monitoring owing to its global coverage and all-weather capability. Synthetic Aperture Radar (SAR) altimetry has become an important tool for coastal sea-level monitoring due to its enhanced along-track resolution (~300 m). However, the large cross-track footprint (~15 km) allows land contamination and complex nearshore scattering to distort radar echoes, leading to significant deviations from standard ocean waveforms. Waveform retracking is an effective approach for mitigating these effects. Nevertheless, existing retrackers typically rely on independent single-waveform or 20 Hz along-track processing, which has limited effectiveness under complex coastal conditions. Furthermore, commonly used retrackers often require leading-edge fitting to obtain initial significant wave height (SWH) estimates, which is sensitive to leading-edge distortions. As a result, sea surface height (SSH) measurements within 5 km of the coastline, especially in the nearshore zone within 3 km, suffer from low data availability and reduced precision. To address these limitations, we propose a spatiotemporal continuity-constrained multi-parameter sub-waveform retracker (MulPOS-C), based on the multi-parameter optimized sub-waveform (MulPOS) determination framework. MulPOS-C organizes repeated-cycle altimetric observations into a two-dimensional (2D) grid (along-track bin × cycle) and identifies optimal sub-waveforms independently along the spatial and temporal dimensions. A spatiotemporal continuity function is then introduced to reconcile directional discrepancies, yielding a consistent set of optimal sub-waveforms across the 2D domain. Sea surface height (SSH) is subsequently retrieved using a fixed power threshold to define the start and end gates of the water-related sub-waveform. This strategy ensures physical consistency and controllable errors relative to full-waveform physical retrackers, while avoiding the instability associated with initial significant wave height (SWH) estimation. Global validation against four retrackers demonstrates that MulPOS-C substantially improves both data availability and precision. Within 1 km of the coastline, 81.8% of stations achieve the lowest RMSE among all methods. At 1 km offshore, the RMSE and SSH noise are reduced to 9.4 cm and 10.6 cm, respectively, representing reductions of at least 4.3 cm in RMSE and 3.5 cm in SSH noise compared with the four retrackers. At 2 km offshore, the RMSE decreases to 7.7 cm and the proportion of gross errors is reduced to 1.8%, significantly outperforming the other retrackers, with all metrics approaching open-sea levels. At 3 km offshore, all metrics reach open-sea levels. In addition, MulPOS-C is methodologically generic and can be readily extended to conventional Low-Resolution Mode (LRM) altimeter data.