PEMODELAN MEAN SEA LEVEL (MSL) DI KOTA SEMARANG DENGAN PENDEKATAN REGRESI NONPARAMETRIK DERET FOURIER

Abstract

The statistical method used to estimate or estimate sea level is by nonparametric regression approach of Fourier series. The problem of flooding due to rising sea levels in Semarang includes problems that have not been solved yet. This resulted in the need for modeling to predict and find out how high the average rising sea level. Fourier series have a fluctuative data pattern due to its periodic nature. This makes the Fourier series as an appropriate approach for modeling the mean sea level or called the Mean Sea Level (MSL).
Before modeling the MSL data with fourier approximation approach, first determine the optimal K value, based on optimal K determination with GCV method obtained K = 277. The result of MSL modeling on tide data of Semarang City with Nonparametric Regression approach Fourier R2 obtained R2 of 95% and MSE = 4,42. Maximum MSL modeling results or average sea level experienced maximum tides occurred on 31 August 2016 and minimum MSL or so-called mean sea level experienced minimum receding occurred on March 2, 2016.
Keywords: MSL, Nonparametric Regression, Fourier Series