Tide modelling using the Kalman filter

Abstract

Away from the conventional Least Squares method of predicting tide, this research explores the use of the Kalman Filter. This work is based on seven harmonic constituents: M₂, S₂, N₂, K₂, K₁, O₁ and P₁. The computed tidal form factor F = 0.1955 indicates that the tide is semi-diurnal, which is an indication that the result of the experiment can be trusted, since Nigerian coastal waters are characteristically semi-diurnal in nature. The computed r² for the Kalman Filter and Least Squares are r² = 0.9708 and r² = 0.9668, respectively. This result indicates that the result of the Kalman Filter is slightly better than the Least Squares.