A correction method of the model output by measurements based on diffusion approximation and its application to the analysis of hydro-physical characteristics

A new data assimilation method or a method of correction of the model outputs by observed data is proposed. The method is based on the diffusion approximation of the optimal filter and it has a number of advantages in comparison with existed data assimilation methods. In particular the method does not need the linearity of model operator, it reduces the problem of optimal filtering to the solution of a system of linear equations, and it allows to constrain the tolerance bounds for the optimal estimator. The conditions of existence and uniqueness of the sough estimation (filter) are found out in the paper. The experiments with the hydrodynamic model НОРЕ (Hamburg Ocean Primitive Equation) are presented and analyzed.