FOURIER SERIES NONPARAMETRIC REGRESSION FOR THE MODELIZING OF THE TIDAL

Abstract

The method of statistic used to estimate the estimation of sea water level is by nonparametric regression approaching of Fourier series. The rob flood caused by sea level rise in Semarang becomes a dissolved problem until today This results the need of modeling to predict and know how high sea level is.
The fourier series have fluctuative data pattern because of its periodic character. This makes Fourier series as the appropriate approaching to modelize the sea tidal. Before modelizing the sea tidal with Fourier series approaching, It is previously necessary to find the optimal K value . Based on the determination of optimal K value, with GCV method, It is obtanied K equals 277. The result of average data of the Semarang sea tidal with reggression nonparametic method showed that R 2 is 95% and MSE = 4,42. The lowest tidal
estimation resulted in Semarang is on March 2, 2016. Then the highest tidal estimation in Semarang City
occurred on August 31, 2016.
Keywords : Nonparametric Regression, Fourier Series, Tidal Sea