Answers are grounded in this paper and cite their sources.
Data assimilation improves the accuracy of numerical models by integrating observational data into model predictions, with the ensemble Kalman filter (EnKF) being a widely used technique. Despite its effectiveness, EnKF is often computationally demanding, which limits its practical use in large-scale, high-resolution, and real-time operational models. To overcome this limitation, the present study applied an alternative approach: optimal interpolation (OI). In OI, the Kalman gain was parametrized and precomputed based on prior model statistics, eliminating the need for extensive ensemble simulations. As a result, the computational cost of a data assimilation run becomes comparable to that of a stand-alone hydrodynamic simulation, making it suitable for real-world applications. This method was implemented in TELEMAC-3D, a three-dimensional hydrodynamic modeling software based on the shallow-water equations, and tested in the Scheldt Estuary, where measured salinity data were assimilated into an operational forecasting system supporting a tunnel immersion project near Antwerp of Belgium. The results show that the model maintains stable salinity fields, progressively reduces relative errors over time, and captures the main hydrodynamic conditions in the estuary. These findings demonstrate the potential of OI as a computationally efficient and accurate data assimilation method, suitable for both real-time forecasting and engineering applications in which efficiency and precision are critical.
No takes yet. Share an insight, caveat, or question.
Chu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: