Application of Kalman filter to short-term tide level prediction

Prediction of tide levels is one of the important problems in determining the schedule making of activities or constructions in coastal and oceanic area. Accurate predictions of tide levels could not be obtained without a long length (about one month or more) of tide measurements by conventional methods. A method of using a short length (a few days) of tide measurements to predict tide levels over a longer time is proposed in the present paper. In the method, a harmonic tide level model is used to predict tide levels. The parameters of the tide level model, that is, the amplitudes of the harmonic components, are estimated by Kalman filtering technique using a few-day tide record with the assumption of known angular frequencies. The proposed model is then applied to the short-term tide level prediction of the Kaohsiung Harbor, Taiwan. As a result, the tide levels predicted by the proposed method are in good agreement with the observed data.