tide numerical model water depth estimation method based on ensemble Kalman filtering

Abstract

The invention discloses a tide numerical model water depth estimation method based on ensemble Kalman filtering. The tide numerical model water depth estimation method comprises the following steps: (1) determining ocean numerical model water depth parameters; (2) generating a set and carrying out numerical simulation: superposing an unbiased Gaussian random number on each water depth increment parameter to generate a water depth increment parameter set so as to generate a water depth parameter set, and substituting the water depth parameter set into the ocean numerical mode to carry out free integration until the ocean numerical mode is stable; (3) assimilating by adopting an enhanced parameter correction data assimilation method; (4) after assimilation is finished, freely integrating the ocean numerical mode to be stable by using assimilated parameters so as to obtain an optimized ocean numerical mode state variable; and (5) carrying out harmonic analysis on the ocean numerical mode state variable to obtain an optimized tide harmonic constant analysis result for tide forecasting. According to the method, observation data are assimilated into the tide numerical model through the EAKF, the water depth parameters are optimally estimated, and the tide simulation precision is improved.

Description

technical field [0001] The present invention relates to ocean data assimilation technology, in particular to a tide numerical model water depth estimation method based on ensemble Kalman filter (Ensemble adjustment Kalman filter, EAKF), which is mainly applied to ocean tide numerical simulation and forecast. Background technique [0002] With the continuous development of computer technology, the ocean numerical model has increasingly become an important tool for people to study and predict the ocean. For all ocean numerical models, the parameterization scheme and the precise parameter values ​​given have an important impact on the numerical simulation results. In the process of ocean numerical model debugging, the parameter values ​​are usually given by a trial and error method, so that the simulation results are close to the observations. Due to the complexity of ocean numerical models, this is an extremely computationally and human-intensive process. Therefore, one of t...

AI Technical Summary

Problems solved by technology

However, the adjoint method has its limitations: for different modes, the adjoint method needs to write the corresponding adjoint mode to obtain the gradient of the objective function of the mode, so the method has a high workload and poor portability

However, there is no relevant report on the research on water depth estimation in shallow seas on the continental shelf.

Images