A Variational Method for Reconstructing and Separating Balanced Motions and Internal Tides From Wide‐Swath Altimetric Sea Surface Height Observations

Abstract

The surface water and ocean topography (SWOT) satellite mission marks a major advance in observing ocean surface dynamics, providing sea surface height (SSH) measurements at unprecedented spatial resolution. Reconstructing gridded SSH fields from SWOT data requires new algorithms capable of retrieving measured fine‐scale variability, which is primarily governed by the coexistence of two dynamical processes: balanced motions and internal tides. As they affect ocean surface currents differently, these processes must be separated in the reconstruction to avoid misinterpretation. A variational data assimilation method is proposed for dynamically reconstructing and separating balanced motions and internal tides from altimeter SSH observations including SWOT. The approach minimizes a cost function involving two distinct models: a quasi‐geostrophic model for balanced motions and a linearized shallow‐water model for internal tides, each controlled by separate sets of variables projected onto adapted reduced bases. The method is implemented and evaluated in an Observing System Simulation Experiment framework over a region centered around Hawai'i which is characterized by strong internal tide and balanced dynamics. The results demonstrate that the method successfully separates the two dynamical contributions and reconstructs a large portion of their variability. Sensitivity analyses show that explicitly estimating the internal tide notably improves the reconstruction of balanced motions. The method also shows potential for recovering part of the non‐stationary (incoherent) internal tide signal.