Attempt for accuracy improvement of wave hindcasting / forecastig in coastal sea and inner sea by data assimilation

Two different types of data assimilation methods, the optimum interpolation method and the adjoint method, are applied to improve wave hindcasting / forecasting accuracy in coastal sea and inner sea of Japan. The optimum interpolation method is a practical and convenient one for improving estimation accuracy of waves in coastal sea facing open sea where there are observation data around the sea. On the other hand, wave characteristics in inner sea are local, being strongly influenced by the surrounding topographic conditions. In such sea area, the adjoint method is an alternative method to improve wave estimation accuracy, in which model parameters are turned to adapt to the local conditions in the area.