Prediction of Tide level by using Holtz-Winters Exponential Smoothing: Case study in Cilacap Bay

Abstract

Sea level rise is a phenomenon that causes the sea level to rise to some extent, And the impact from The changes in the tide level can influence flooding in coastal area that can damage the structure of the building around that area and also disturb the health of functionally linked neighboring ecosystems. But now with the development of technology and science it is possible to make some projection of the future tidal level by using a time series data, which is very important for an island country like Indonesia, this forecasted data can be used to make a planning and implementing a projects in port and coastal area. Now, there are many methods to predict the future value of several things. In this paper, the Holt-Winters Exponential Smoothings applied to forecast the tidal level in Cilacap. Then the Holt-Winters forecasting performance compared with the Autoregressive Integrated Moving Average (ARIMA), and Seasonal-Autoregressive Integrated Moving Average (SARIMA), in order to see which one that can produce the best forecast. The method performance measured by using root mean square error (RMSE) and R-Square. The Holt-Winters Exponential smoothing produces RMSE and R-Square that are better than ARIMA and SARIMA. The choice of seasonal period significantly affects the forecasting result produced by the Holt-Winters method.