WADDMD: A Multiframe Mixed-Phase Decoupling Algorithm for GEOSC-SAR Sea Surface Wave Multidimensional Parameter Retrieval

Accurate retrieval of multidimensional sea surface parameters, including height, radial velocity, and their temporal evolution, is critical for ocean monitoring and environmental applications. Dual-satellite geosynchronous spatial correlation synthetic aperture radar (GEOSC-SAR) configurations, benefiting from the persistent monitoring of geosynchronous orbits, enable multiframe interferometric observations that capture rich information on the evolution of dynamic sea surfaces. However, their cross- and along-track components inherently form hybrid baselines, yielding mixed interferometric phases in which height and radial velocity are coupled. Existing decoupling methods typically require additional baselines or revisit passes, increasing system complexity. This letter presents a wave-adaptive dynamic decomposition and mixed-basis decoupling (WADDMD), a novel algorithm that resolves mixed-phase coupling using only a single hybrid baseline. WADDMD integrates the following: 1) wave-adaptive dynamic decomposition, which embeds linear wave theory into dynamic mode decomposition to construct physically consistent temporal bases and 2) mixed-basis decoupling, which structurally separates height and velocity contributions via baseline sensitivity coefficients and regularized least-squares fitting. Extensive validation confirms the method’s effectiveness: realistic GEO-SAR simulation experiments demonstrate significant improvement over direct inversion; evaluation with in situ buoy observations further verifies accurate retrieval of sea surface height, radial velocity, and temporal evolution; and multiregion tests using Copernicus Marine Service data validate its robustness across different oceanic regions.