Data,Figure,Code.zip

In this study, an ensemble adjustment Kalman filter (EAKF) based full-grid parameter optimization scheme, namely the EAKF-smoothing (EAKF-S) scheme, is proposed. By smoothing the ensemble members of posterior parameters, full-grid parameter optimization is successfully realized. A Bohai Sea and Yellow Sea tide model (BYM) was built, which models the 8 principal tidal constituents simultaneously. Twin and practical experiments of bathymetry estimation were conducted based on water level observations. By applying the EAKF-S, forecasting accuracy of the BYM model was significantly improved. The effectiveness and robustness of the EAKF-S method were confirmed. This study provides a practical solution for the optimization of geographic parameters, benefit to determining spatial varying parameters in ocean modelling.